| *** camunoz has quit IRC | 00:00 | |
| *** prad has quit IRC | 00:00 | |
| clarkb | jogo: I do not see a REPO_ROOT there | 00:00 |
|---|---|---|
| jogo | clarkb: http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/hacking.yaml#n10 | 00:01 |
| clarkb | oh I must be behind | 00:01 |
| jogo | clarkb: I update my local copies on a regular basis with https://github.com/jogo/mustached-bear | 00:02 |
| clarkb | jogo: if we can get my zuul-cloner change in we can switch to using that (I guess I need to fix the bug there too) | 00:02 |
| clarkb | jogo: you should use zuul-cloner :) | 00:02 |
| * jogo looks at zuul-cloner | 00:02 | |
| clarkb | jogo: its missing a change to make it useful locally though, but I pushed it up so hopefully soon | 00:02 |
| *** gordc has quit IRC | 00:03 | |
| greghaynes | clarkb: you have a -1 on your zuul cloner patch fyi | 00:04 |
| jogo | clarkb: where does it get a list of repos to clone from? | 00:04 |
| clarkb | greghaynes: gah thanks | 00:04 |
| clarkb | jogo: you provide it on the command line or in a "map" file that maps sources to dests | 00:04 |
| greghaynes | jogo: http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/diskimage-builder.yaml is an example | 00:05 |
| *** sputnik13 has quit IRC | 00:05 | |
| *** dannywilson has quit IRC | 00:05 | |
| clarkb | there is an issue with our git cache repos interacting with GitPython that I need to sort out too | 00:05 |
| jogo | ahh | 00:05 |
| openstackgerrit | John Griffith proposed openstack-infra/project-config: Add config for Cinder LVM Thin configuration https://review.openstack.org/182499 | 00:05 |
| *** hdd_ has joined #openstack-infra | 00:05 | |
| jgriffith | bahhh | 00:06 |
| jogo | I just fetch/clone/pull everything except the -attic orgs | 00:06 |
| jgriffith | zuul piece | 00:06 |
| *** markvoelker has joined #openstack-infra | 00:06 | |
| *** achanda_ has joined #openstack-infra | 00:06 | |
| clarkb | jogo: then zuul-cloner will work well for you | 00:07 |
| jogo | and dhellmann has a bash script that does the same thing somewhere too | 00:07 |
| clarkb | jogo: you point it at that cache and it makes working repos for you that are correct based on zuul parameters or it falls back safely to sane defaults | 00:07 |
| greghaynes | jogo: I think theres like 5 tools around openstack that do this :( | 00:07 |
| clarkb | jogo: the real killer feature is you should be able to use the same tool in the gate to get the right thing to test and locally to do dev | 00:07 |
| clarkb | greghaynes: and zuul-cloner is the only one you can use in the gate | 00:08 |
| clarkb | greghaynes: which is wh it wins imo | 00:08 |
| jogo | clarkb: well I just used it to keep my repos fresh so grep works accross them | 00:08 |
| jeblair | dougwig, fungi: is someone working on making the neutron jobs use zuul-cloner? | 00:08 |
| jogo | I query gerrit for the API list and go | 00:08 |
| openstackgerrit | Clark Boylan proposed openstack-infra/zuul: Simplify cloner required parameters https://review.openstack.org/181694 | 00:08 |
| clarkb | greghaynes: ^ thanks | 00:08 |
| clarkb | jeblair: we need 181694 first, so I guess I am | 00:09 |
| greghaynes | We also should fix that bug first | 00:09 |
| clarkb | greghaynes: ++ | 00:09 |
| greghaynes | its not terribly useful otherwise | 00:09 |
| jogo | clarkb: but my thing is more of just a fun little script to mess around with then anything else | 00:09 |
| jeblair | clarkb: why is 181694 needed? | 00:09 |
| openstackgerrit | Merged openstack-dev/hacking: Don't shallow clone from file:// in integration test https://review.openstack.org/182494 | 00:09 |
| dougwig | jeblair: no, i was going to stop force cloning in the CI. but can you point me at what youre' talking about? | 00:09 |
| clarkb | jeblair: because otherwise zuul-cloner does not work without a zuul env and these scripts are expected to run locally on `tox` | 00:09 |
| *** achanda has quit IRC | 00:10 | |
| jeblair | clarkb: oh, well, i'm interested in neutron using zuul-cloner in the gate | 00:10 |
| clarkb | me too | 00:10 |
| clarkb | but one of their requirements is it work locally too | 00:10 |
| jeblair | oh i see | 00:10 |
| clarkb | 181694 should allow that | 00:10 |
| jeblair | dougwig: zuul-cloner is what every job should (and will in the not too distant future) use to set up the git repos under test | 00:11 |
| *** achanda_ has quit IRC | 00:11 | |
| *** banix has quit IRC | 00:11 | |
| jeblair | clarkb: so assuming your patch lands, how will the various neutron repos be set up and installed for the test? | 00:11 |
| jeblair | (eg: zuul runs tox, and tox runs zuul-cloner, and...?) | 00:12 |
| *** camunoz has joined #openstack-infra | 00:12 | |
| clarkb | yup and each tox target could have a specific set of repos to update | 00:12 |
| *** derekh has quit IRC | 00:12 | |
| jeblair | but that means that each repo has to maintain tox targets in sync... | 00:13 |
| clarkb | because right now its all adhoc, hacking does one thing, neutron another | 00:13 |
| openstackgerrit | John Griffith proposed openstack-infra/project-config: Add config for Cinder LVM Thin configuration https://review.openstack.org/182499 | 00:13 |
| jeblair | clarkb: how is hacking involved? | 00:13 |
| clarkb | jeblair there should only be one for each test | 00:13 |
| jeblair | clarkb: but we're talking about integration tests between neutron repos and neutron, right? | 00:13 |
| clarkb | yes but only one repo has the tox target | 00:14 |
| jeblair | gotach | 00:14 |
| clarkb | hacking is involved because it also breaks integration testing like neutron | 00:14 |
| clarkb | jeblair: https://review.openstack.org/#/c/182494/ | 00:14 |
| clarkb | it fixes a regression so I approved it but that script has the same general problem | 00:14 |
| jeblair | okay, so neutron and neutron-* all run a job that zuul-cloners neutron, and then runs "tox something" which zuul-cloners neutron and all the plugins and then installs them? | 00:15 |
| *** salv-orlando has joined #openstack-infra | 00:15 | |
| jeblair | (so all the plugins and neutron end up running a tox target from neutron) | 00:15 |
| *** esker has joined #openstack-infra | 00:16 | |
| clarkb | ya thats the idea | 00:16 |
| clarkb | basically each integration test has to have one source for the test itself, but then we can wrap zuul-cloner around that as necessary to get the right refs for all the deps | 00:16 |
| jeblair | we can do that with only one zuul-cloner call, but i guess it won't hurt either way | 00:17 |
| jeblair | (if we specify the repos under test in the shared job definition) | 00:17 |
| *** annegentle has joined #openstack-infra | 00:17 | |
| clarkb | a more concrete example is change to networking-foo comes in, it runs job that grabs latest neutron and networking-foo with zuul-cloner then it runs tox in neutron | 00:17 |
| clarkb | jeblair: ya | 00:17 |
| jeblair | clarkb: okay, i think we're on the same page; dougwig do you follow? | 00:18 |
| jeblair | clarkb: so how is hacking involved? not actually in this scenario, right -- just a similar problem? | 00:18 |
| clarkb | or I guess to make it more local dev friendly you might want to run tox and have it do a second repo update just so that you don't always have to start from latest neutron? | 00:19 |
| clarkb | jeblair: correct similar problem | 00:19 |
| *** puranamr has quit IRC | 00:19 | |
| clarkb | jeblair: hacking has a job that tests against nova as an integration test. Except that test won't properly apply zuul ref'd nova for hacking if I am reading correctly | 00:19 |
| *** banix has joined #openstack-infra | 00:19 | |
| clarkb | so its possible to wedge things by getting a change in at the right moment | 00:19 |
| *** puranamr has joined #openstack-infra | 00:20 | |
| jeblair | clarkb: does nova use CD hacking? | 00:20 |
| jeblair | clarkb: i thought nova would use a hacking release | 00:20 |
| clarkb | jeblair: no, it uses a release | 00:20 |
| *** chlong_ has joined #openstack-infra | 00:20 | |
| clarkb | jeblair: we would wedge hacking dev not nova's | 00:20 |
| clarkb | but maybe thats ok | 00:20 |
| jeblair | clarkb: yeah, but that can happen anyway | 00:20 |
| *** otter768 has joined #openstack-infra | 00:21 | |
| jeblair | clarkb: if no hacking changes are in flight, you can still land a change that will wedge hacking | 00:21 |
| jeblair | clarkb: if no hacking changes are in flight, you can still land a change to nova that will wedge hacking | 00:21 |
| clarkb | ya | 00:21 |
| jogo | nothing CDs hacking | 00:21 |
| jeblair | so i think in this case, we're just wanting to use zuul-cloner to do the right thing with git repos | 00:21 |
| jogo | the hacking integration check that runs on the nova tree, is a non-voting job to get some information on what a test looks like on a bigrepo | 00:22 |
| jeblair | we won't solve that particular wedge case with zuul-cloner, we either need to make that job non-voting, co-gate with nova, or accept that the current risk profile is the one we want | 00:22 |
| jeblair | oh okay, non-voting sounds like exactly the right thing for that | 00:22 |
| jogo | also it always returns success | 00:22 |
| clarkb | ah ok | 00:22 |
| jeblair | jogo: why always return success if it is non-voting? | 00:23 |
| *** annegentle has quit IRC | 00:23 | |
| jogo | jeblair: because it runs hacking with all tests enabled | 00:23 |
| jogo | it would always fail otherwise | 00:23 |
| jogo | I haven't been able to come up with a useful binary value to return | 00:23 |
| jeblair | okay, so you're reading the output. at least, theoretically. :) | 00:24 |
| jogo | jeblair: correct | 00:24 |
| jogo | when looking at reviews for new rules I often do | 00:24 |
| *** puranamr has quit IRC | 00:24 | |
| jeblair | makes sense | 00:25 |
| *** otter768 has quit IRC | 00:25 | |
| *** salv-orlando has quit IRC | 00:26 | |
| jogo | jeblair: here is a nice example http://logs.openstack.org/94/157894/3/check/gate-hacking-integration-nova/0d6fa16/console.html.gz#_2015-02-28_12_27_21_285 | 00:26 |
| *** mtanino has quit IRC | 00:27 | |
| *** bhuvan is now known as bhuvan-afk | 00:28 | |
| *** david-lyle has quit IRC | 00:28 | |
| jeblair | clarkb: i think there's a fundamental problem in your change | 00:29 |
| clarkb | there may be | 00:30 |
| jeblair | clarkb: because of the mechanism for pulling vars from the environment is to set the command line parameter defaults for each of the ones in the list to what's in the environment | 00:31 |
| *** annegentle has joined #openstack-infra | 00:33 | |
| *** esker has quit IRC | 00:34 | |
| clarkb | jeblaur it does check all are set if any are set | 00:34 |
| dougwig | jeblair: i believe so, and i don't think i need to mess with it, correct? i just need to trust that the local repos have been setup correctly, and not reach around them. | 00:34 |
| jeblair | dougwig: that's the idea. zuul-cloner has one job and we expect it to do it well. :) | 00:35 |
| jeblair | clarkb: i think removing something from the list will cause it to stop being a parameter at all. the parameters are built from that list. | 00:36 |
| jeblair | clarkb: (ZUUL_ENV_SUFFIXES) | 00:36 |
| clarkb | jeblair correct so the ones I removes are unused | 00:36 |
| *** shashankhegde has quit IRC | 00:37 | |
| jeblair | clarkb: oh, you are saying zuul closer just flat does not actually use them? | 00:37 |
| clarkb | ya | 00:37 |
| jeblair | clarkb: ah, then i think your patch may be fully correct. :) | 00:38 |
| jeblair | clarkb, dougwig: so what will actuall install neutron and the plugins inside the tox environment? | 00:40 |
| *** zz_dimtruck is now known as dimtruck | 00:40 | |
| jeblair | clarkb, dougwig: will we specify the dependencies in tox or add them to the neutron requirements file, or something else? | 00:41 |
| dougwig | jeblair: what i'd like to try is lifeless's suggestion of putting neutron into requirements, making the local unit job install it manually (as now), and all other tox jobs just use pip, which should grab it local (plus or minus some command-line tweaks). but... since neutron isn't in global-requirements, either crossing that bridge first, or adding | 00:43 |
| dougwig | something like project-requirements.txt that tox can use as an additional requirements file. | 00:43 |
| dougwig | that change would be in the sub-repos. neutron itself would not change. | 00:43 |
| dougwig | that's not without its own warts, since we can't version pin at all, but in CI land, that shouldn't be an issue. | 00:44 |
| jeblair | dougwig: if you want to run the same job on multiple repos, then you should define that job in one place. i had been assuming you would define it as a tox env on neutron. where do you want to do define it? | 00:44 |
| *** ddieterly has joined #openstack-infra | 00:44 | |
| dougwig | it's an lbaas api job, defined in the lbaas repo. it then runs against lbaas changes or neutron changes. | 00:45 |
| dougwig | right now, it's a tox env in lbaas. | 00:45 |
| jeblair | dougwig: are the only repos involved neutron and neutron-lbaas? | 00:45 |
| dougwig | jeblair: correct. | 00:45 |
| *** marun has quit IRC | 00:48 | |
| anteaya | dougwig: this is the same pattern for fwaas and vpnaas too, yes? | 00:48 |
| dougwig | no, fwaas has no co-gates, and vpnaas doesn't use tox in its single co-gate, so it doesn't run afoul of this. but i will modify them to use whatever mechanism lbaas ends up adopting for running unit tests outside of CI. | 00:50 |
| jeblair | okay. that makes sense to have the tox env on the lbaas repo do the setup for this. but the details about how to actually install it are important. i would lean towards having the tox env specify that they should be installed from pre-existing local repos. that could be done now without clarkb's patch, but it will make it harder (but not impossible) to use that repo for local devolpment. | 00:50 |
| dougwig | (previous was for anteaya) | 00:50 |
| openstackgerrit | Tony Breeds proposed openstack-infra/irc-meetings: Remove Blazar Meeting https://review.openstack.org/182509 | 00:50 |
| openstackgerrit | Tony Breeds proposed openstack-infra/irc-meetings: Move Cue meeting to Monday https://review.openstack.org/182510 | 00:50 |
| *** aduarte has joined #openstack-infra | 00:51 | |
| dougwig | jeblair: i have to run out. can i add you as a reviewer and we can refine there? | 00:51 |
| jeblair | dougwig: please, but i'm also deferring to clarkb since it sounds like he has a plan :) | 00:51 |
| jeblair | i'm just trying to help out | 00:52 |
| dougwig | jeblair: ok, i'll add him too. :) | 00:52 |
| dougwig | thank you, i appreciate it. | 00:52 |
| clarkb | well semi plan mostly trying to get beaker working too and its similar issues | 00:52 |
| *** ddieterl_ has joined #openstack-infra | 00:52 | |
| jeblair | clarkb: the disconnect for me is that our ci system can know where to put the repos; i'm not sure where a developer-run test should place git checkouts | 00:53 |
| jeblair | splatting them in ~, for instance, may not be the right answer. | 00:53 |
| clarkb | ya, for beaker it can know because its a vm but good point about neutron | 00:53 |
| dougwig | jeblair: if that's really a holdup, i'd be fine with one env for devs and one for ci. | 00:53 |
| *** rwsu has quit IRC | 00:54 | |
| *** ddieterly has quit IRC | 00:54 | |
| *** stevemar has joined #openstack-infra | 00:54 | |
| jeblair | well, i mean, we could decide that they go in .tox/git/ or something | 00:54 |
| dougwig | but i think lifeless' point was that if they're running it local, they must already have devstack, so neutron is going to be there. | 00:54 |
| openstackgerrit | John Griffith proposed openstack-infra/project-config: Add config for Cinder LVM Thin configuration https://review.openstack.org/182499 | 00:55 |
| jeblair | but tox creates a virtualenv so if devs are running tox, whatever devstack installed will be invisible to it | 00:55 |
| dougwig | but the pip that runs outside to venv, to set it up, can see the local package. so if it's in requirements (unpinned), it'll just reuse it. right? | 00:56 |
| dougwig | /to venv/the venv/ | 00:56 |
| jeblair | only if you set tox to re-use site packages, which we don't, because it always messes something up | 00:57 |
| dougwig | oy. :) | 00:57 |
| *** doug-fish has left #openstack-infra | 00:58 | |
| dougwig | ok, i need to eat, and i'm really high on cold meds right now. which is either making pip easier to understand or harder. maybe i can selectively use the site package for neutron only? | 00:58 |
| dougwig | jeblair: ^^ | 00:58 |
| anteaya | dougwig: ha ha ha | 00:58 |
| jeblair | dougwig: no, it's all or none: http://tox.readthedocs.org/en/latest/config.html#confval-sitepackages=True|False | 00:58 |
| jeblair | dougwig: and i can't convey the amount of pain that ensues whenever we have set that to true in the past :) | 00:59 |
| *** ayoung has joined #openstack-infra | 00:59 | |
| dougwig | jeblair: not arguing to go global. :) but i can pass any flag i want to the deps, not just -r. | 00:59 |
| lifeless | jeblair: ah, i thought we had site-packages on on most venvs? | 00:59 |
| lifeless | dougwig: so, are you ok with forcing devs to pull down a new neutron? | 01:00 |
| *** alexsyip has quit IRC | 01:00 | |
| jeblair | lifeless: i think it's the opposite. i think we might even have finally gotten it to be disabled everywhere. | 01:01 |
| *** esker has joined #openstack-infra | 01:01 | |
| dougwig | lifeless: i'd prefer not, but i was going to get around that by proposing that we force a git clone if we don't see CI env vars, along with a TODO on that line to remove it one the server package hit pypi. | 01:01 |
| lifeless | jeblair: cool | 01:01 |
| lifeless | jeblair: I heartily support such endeavours | 01:01 |
| jeblair | yes, so much pain | 01:01 |
| *** yamamoto has joined #openstack-infra | 01:01 | |
| dougwig | /one/once/ | 01:01 |
| lifeless | dougwig: well as I keep coming back to, we have to install non-pypi for CI | 01:01 |
| dougwig | " if we don't see CI env vars" | 01:01 |
| lifeless | dougwig: so the ting being on pypi is not relevant for the CI discussion. It may be relevant for the dev discussion, but only if lbaas can run with 'last release' of neutron | 01:02 |
| dougwig | my comment was solely confined to the dev discussion. | 01:02 |
| lifeless | dougwig: since i don't know - can lbaas run with a previous release of neutron? | 01:02 |
| dougwig | no, it certainly can not in all cases (and this problem spreads to about 20 repos, not just lbaas.) i'm operating under the assumption that we can release for each of the branches regularly, as if it was a library (and if that means 'experimental' flag for master, fine.) | 01:04 |
| dougwig | lifeless: ^^ | 01:04 |
| dougwig | but, wait, back to your original question, we can auto-install neutron for devs in the non-CI case until those details are sorted out, as today. | 01:05 |
| *** bhuvan-afk has quit IRC | 01:06 | |
| lifeless | dougwig: well, what it means is that for the dev story we want git neutron, and for CI we want git neutron. | 01:06 |
| *** bknudson has joined #openstack-infra | 01:07 | |
| lifeless | dougwig: My personal inclination would be to do two things; set the source dir in pip.conf in CI to point to the checked out neutron (solves the cloning-fresh issue) | 01:07 |
| dougwig | lifeless: i failed to parse that. it means that for the dev story we want (pip install git+url), and for the CI we want (pip install local). | 01:07 |
| lifeless | secondly change install_command in tox to include the git neutron with ZUUL refs (or whatever jeblair recommends) | 01:08 |
| lifeless | dougwig: no - pip install git+url both times; tell pip where the local git tree is for CI | 01:08 |
| jeblair | lifeless: can you elaborate on "pip.conf in CI to point to the checked out neutron" ? | 01:08 |
| jeblair | lifeless: not sure what that means | 01:09 |
| *** sarob has quit IRC | 01:09 | |
| lifeless | jeblair: for git+ urls pip will check them out to a source directory. It defaults to src at the root of the venv. | 01:09 |
| lifeless | jeblair: but we can set it in ~/.pip/pip.conf - if we set that to /opt/stack/new, it will find the existing neutron git tree | 01:09 |
| *** ivar-laz_ has joined #openstack-infra | 01:09 | |
| *** baoli has quit IRC | 01:10 | |
| *** ivar-lazzaro has quit IRC | 01:10 | |
| dougwig | lifeless: oh, that's pretty neat. | 01:10 |
| lifeless | jeblair: if you look in pip install --help at the option --src | 01:10 |
| jeblair | lifeless: ok. i've never used a git+ url with a zuul checkout before, and i'm wary of it. the pattern that i prefer is that something (zuul-cloner) set up git repos, and then something install from those repos. | 01:10 |
| jeblair | lifeless: basically, i don't think anything should perform git operations except zuul-cloner | 01:11 |
| lifeless | ok, you can certainly do that too | 01:11 |
| lifeless | as long as we edit the requirements before pip encounters them | 01:11 |
| jeblair | that's not ideal either | 01:11 |
| *** signed8bit_ZZZzz is now known as signed8b_ | 01:12 | |
| lifeless | here are the constraints (in current pip) | 01:12 |
| lifeless | 1) you can't give the same requirement twice in different ways. You *might* be able to give the exact same one twice, but I'd need to check that. | 01:12 |
| *** claudiub|2 has quit IRC | 01:12 | |
| lifeless | 2) whats in the repository should not reference locations, because they can't be reflected sensibly into setuptools kwargs. | 01:13 |
| dougwig | i have to run out. i will check back in after dinner, and hopefully see a consensus that i can implement. thanks, all. :) | 01:13 |
| lifeless | (And we just removed the local glue that kindof-sometimes-made-that-work) | 01:13 |
| anteaya | thanks dougwig | 01:13 |
| jeblair | lifeless: but we can have requirements pre-installed, possibly by tox. | 01:14 |
| lifeless | 3) we're running under tox, so there's no convenient point to stop and install things | 01:14 |
| lifeless | jeblair: hah :) | 01:14 |
| lifeless | jeblair: there was a thread about this last week on -dev IIRC. | 01:14 |
| lifeless | we can hack install_command to do something | 01:14 |
| jeblair | lifeless: we can literally specify requirements in tox | 01:14 |
| lifeless | jeblair: yes | 01:14 |
| jeblair | we use it all the time | 01:14 |
| jeblair | deps= | 01:15 |
| lifeless | jeblair: its my understanding that tox needs to work for both devs and users | 01:15 |
| lifeless | erm | 01:15 |
| lifeless | devs and CI | 01:15 |
| jeblair | lifeless: yeah, i kind of think that we should get it working for ci, and if there's a way to make it work for devs, fine. | 01:15 |
| lifeless | sorry, local interrupt, back soon | 01:15 |
| jeblair | i think dougwig is okay with that | 01:15 |
| jeblair | but we have the pattern _well established_ for ci, so we should implement that here, and think about improving it later | 01:16 |
| jeblair | and the pattern is: checkout, install, run. | 01:16 |
| *** ianw has quit IRC | 01:16 | |
| *** ianw has joined #openstack-infra | 01:16 | |
| dougwig | If it comes to it, I'll burn two tox envs. Worse has been done. | 01:17 |
| jeblair | indeed, much worse has been done :) | 01:17 |
| *** esker has quit IRC | 01:18 | |
| jeblair | dougwig, lifeless: install_command may be the way to make it work for devs and ci, because honestly, i think we need some kind of logic switch like dougwig was talking about | 01:19 |
| jeblair | anyway, my turn to leave. :) | 01:19 |
| lifeless | back | 01:20 |
| anteaya | enjoy your evening | 01:20 |
| lifeless | kk | 01:20 |
| *** btully has joined #openstack-infra | 01:21 | |
| openstackgerrit | Tony Breeds proposed openstack-infra/irc-meetings: Update Neutron Service chaining Meeting https://review.openstack.org/182516 | 01:22 |
| *** marun has joined #openstack-infra | 01:24 | |
| *** claudiub has joined #openstack-infra | 01:28 | |
| fungi | jeblair: clarkb: sorry, was afk. skimmed... the neutron-lbaas api job currently looks like it relies on devstack-gate and devstack to set up neutron, then runs a "tempest" tox env to perform the testing, and as a side effect of initializing that env the neutron-lbaas requirement on neutron getting installed into tempest's virtualenv is merely a side effect | 01:29 |
| fungi | at least that's what i was able to figure out by digging in the devstack-gate hook scripts for it | 01:30 |
| fungi | i suppose i could be misreading them | 01:30 |
| *** Sukhdev has joined #openstack-infra | 01:33 | |
| fungi | the gate hook configures the services overrides for devstack and then the post-test hook copies a custom tempest config into place and runs it under tox like a typical devstack+tempest job | 01:34 |
| fungi | so i _think_ that implementing this with zuul-cloner is basically the "reimplement devstack-gate using zuul-cloner" project? | 01:35 |
| *** zhiwei has joined #openstack-infra | 01:35 | |
| fungi | that "tempest" tox env is simply inheriting the install-this-project's-deps configuration from the default testenv in tox.ini | 01:36 |
| lifeless | sounds like not using a venv would be a win | 01:37 |
| lifeless | if everything is already installed | 01:37 |
| lifeless | can we tell tox topo ? | 01:37 |
| lifeless | e.g. just call testr ? whats tox doing for us in this context? | 01:37 |
| fungi | i think it's simply copying the fact that normal devstack+tempest jobs run tempest from a virtualenv | 01:39 |
| lifeless | hi | 01:40 |
| fungi | though it's not just neutron it's having pip install from a git url. it's doing the same to install tempest from git too | 01:40 |
| lifeless | fungi: ok so that AIUI is to decouple tempest from the global requirements | 01:40 |
| lifeless | anyhow | 01:40 |
| lifeless | we have a more pressing issue | 01:40 |
| lifeless | neutron is still bust | 01:40 |
| fungi | oh? | 01:40 |
| lifeless | and grenade is now failing | 01:40 |
| lifeless | http://logs.openstack.org/38/182438/4/check/check-grenade-dsvm-neutron/0f785be/logs/grenade.sh.txt.gz#_2015-05-13_01_21_01_644 | 01:40 |
| lifeless | fungi: well the non-voting was meant to unblock landing the actual fix to neutron | 01:41 |
| lifeless | fungi: which is 182438 | 01:41 |
| lifeless | fungi: but thats failing in grenade on a omg-it-must-be-separate issue ^ | 01:41 |
| TheJulia | Would anyone happen to have handy documentaiton or a script that one can allow someone build a their own shiny new bare-trusty node as nodepool would have it? | 01:43 |
| *** esker has joined #openstack-infra | 01:44 | |
| *** mestery has quit IRC | 01:45 | |
| fungi | lifeless: i see openstackclient barfing on a too-new python-cinderclient... trying to figure out if that's before or after upgrading | 01:45 |
| lifeless | fungi: yeah, wondering if perhaps its time to force-merge 438 | 01:46 |
| lifeless | fungi: since it passes everything else, and *everything* using neutron master is currently broken | 01:46 |
| fungi | trying to figure out if this is new complication (e.g. from the openstackclient version which just got released) or old broken related to this | 01:47 |
| lifeless | I think its new, but can't prove it | 01:48 |
| lifeless | fungi: its got a bug open. its new | 01:49 |
| anteaya | TheJulia: I think we steer folks to disk-image-builder for that | 01:49 |
| lifeless | fungi: [openstack-dev] [release] python-novaclient 2.25.0 | 01:49 |
| lifeless | fungi: in -dev | 01:49 |
| fungi | looks like between 1.0.3 and 1.0.4 a requirements sync capped cinderclient in openstackclient's reqs on stable/kilo | 01:49 |
| lifeless | 'And openstackclient 1.0.4 broke grenade: | 01:49 |
| lifeless | https://bugs.launchpad.net/python-openstackclient/+bug/1454467 | 01:49 |
| openstack | Launchpad bug 1454467 in python-openstackclient "Latest version 1.0.4 doesn't work with master" [Undecided,New] | 01:49 |
| lifeless | I think we need a 1.1 release for trunk and make sure caps are set so its not used in stable/kilo. Or something like that. | 01:49 |
| lifeless | ' | 01:49 |
| lifeless | [have I mentioned I really don't like this stable-branches-of-non-servers thing? ] | 01:50 |
| fungi | "latest" python-openstackclient is 1.2.0 | 01:50 |
| fungi | lifeless: without stable branches how do we backport fixes to older releases contemporary with the api server releases which use those libraries? | 01:51 |
| anteaya | TheJulia: http://git.openstack.org/cgit/openstack/diskimage-builder/tree/README.rst | 01:51 |
| tchaypo | 2m5s just for git review to tell me it can’t rebase because I have unstashed changes. | 01:51 |
| lifeless | fungi: what we used to do was roll forward | 01:51 |
| lifeless | fungi: AIUI we're doing stable branches of these things because API compat is too hard or something. But I'm going to derail us with a rant if I don't focus, so... lets move past this. | 01:52 |
| fungi | lifeless: yep, and then distros disliked that we didn't provide them with backports to the versions contemporary with the stable releases of the servers too | 01:52 |
| *** camunoz has quit IRC | 01:52 | |
| fungi | moving past ;) | 01:52 |
| lifeless | fungi: so ? :) | 01:52 |
| lifeless | fungi: they aren't the ones wearing the engineering cost of the matrix | 01:52 |
| openstackgerrit | James Polley proposed openstack-dev/pbr: WIP: Teach pbr to read extras from setup.cfg https://review.openstack.org/181795 | 01:53 |
| *** Sukhdev_ has joined #openstack-infra | 01:53 | |
| fungi | so, something is pulling in newer cinderclient than what stable/kilo requiements sync'd to openstackclient | 01:53 |
| fungi | checking now to see who's wrong | 01:53 |
| tchaypo | maybe one day my git review will learn about that and return me to the command line | 01:54 |
| anteaya | one day | 01:54 |
| fungi | stable/kilo global requirements also says cinderclient <1.2.0 | 01:54 |
| *** Sukhdev has quit IRC | 01:54 | |
| tchaypo | real4m9.710s | 01:55 |
| fungi | Collecting python-cinderclient>=1.0.6 (from glance-store<0.5.0,>=0.3.0->glance==2015.1.1.dev1) | 01:55 |
| TheJulia | anteaya: Yah, thats what I would figure, just trying to figure out complete weirdness between bare-trusty, and multiple end user installed trusty machines | 01:56 |
| lifeless | fungi: yes but https://github.com/pypa/pip/issues/2687 and https://github.com/pypa/pip/issues/2731 which will fix this | 01:56 |
| fungi | so glance-store isn't updated with global-requirements on stable/kilo in its last release | 01:56 |
| lifeless | fungi: are still open issues | 01:56 |
| lifeless | anything we install from pip will have old requirements by definition | 01:56 |
| lifeless | unless we release the whole strongly connected set each time any of them change | 01:57 |
| fungi | well, ideally their requirements don't arbitrarily change on stable branches | 01:58 |
| anteaya | TheJulia: what weirdness are you witnessing? | 01:58 |
| *** fedexo has joined #openstack-infra | 02:00 | |
| TheJulia | anteaya: libvirt spitting out "error: unknown OS type hvm" when trying to create a domain | 02:00 |
| lifeless | ok, I need to go | 02:01 |
| lifeless | glwt | 02:01 |
| fungi | okay, i think i see the "problem" | 02:01 |
| *** sarob has joined #openstack-infra | 02:02 | |
| *** claudiub has quit IRC | 02:02 | |
| anteaya | TheJulia: any of your searches providing useful results? | 02:02 |
| anteaya | I got two results both solutions look different | 02:02 |
| fungi | openstack/glance_store is not listed in stable/kilo (or master for that matter) requirements projects.txt and so is not getting requirements update proposals | 02:03 |
| *** patrickeast has quit IRC | 02:03 | |
| anteaya | what do you get? | 02:03 |
| *** esker has quit IRC | 02:03 | |
| anteaya | fungi: that would be a problem | 02:03 |
| greghaynes | TheJulia: I get that when booting a kvm guest without kvm enabled/installed | 02:03 |
| *** camunoz has joined #openstack-infra | 02:05 | |
| anteaya | that is consistent with one of the blog posts I found: http://blog.loftninjas.org/2009/05/04/libvirt-unknown-os-type-hvm/ | 02:06 |
| TheJulia | anteaya: basicaly ^^^ except the kvm module was just not loaded. I went ahead and updated my process go ensure that the module was loaded, still no dice, tried other methods. | 02:06 |
| *** annegentle has quit IRC | 02:07 | |
| greghaynes | TheJulia: I dont entirely understand what youre attempting to do? that should be a hypervisor setting... | 02:07 |
| * TheJulia determines sleep and cofee is required in the morning | 02:08 | |
| greghaynes | heh | 02:08 |
| *** shashankhegde has joined #openstack-infra | 02:08 | |
| TheJulia | greghaynes: basically, fire up a vm that I can start/stop via ironic's pxe_ssh driver | 02:08 |
| *** banix has quit IRC | 02:08 | |
| greghaynes | ah! tripleo | 02:08 |
| greghaynes | :p | 02:08 |
| TheJulia | eh, not exactly | 02:08 |
| *** banix has joined #openstack-infra | 02:08 | |
| greghaynes | (had to make the joke) | 02:08 |
| TheJulia | yeah :) | 02:08 |
| greghaynes | so, that seems like a hypervisor issue WRT the guest definition, not an issue with the guest image | 02:09 |
| greghaynes | which makes me thing that the nodepool image shouldnt be your issue | 02:10 |
| TheJulia | that is what I'm thinking, but i wanted to spin up a replica locally so I could peer into it and compare/contrast | 02:12 |
| *** tmcpeak has quit IRC | 02:12 | |
| *** ihrachyshka has quit IRC | 02:13 | |
| greghaynes | TheJulia: on hpcloud I believe the devstack-trusty images are made with DIB | 02:13 |
| greghaynes | http://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/templates/nodepool/nodepool.yaml.erb#n975 | 02:14 |
| greghaynes | TheJulia: So you need to dib with those elements which come from project-config/nodepool/elements | 02:14 |
| *** ddieterl_ has quit IRC | 02:15 | |
| *** ddieterly has joined #openstack-infra | 02:15 | |
| greghaynes | as for the snapshots (which is how rax works) I have no idea | 02:16 |
| *** shardy_z has quit IRC | 02:16 | |
| *** tiswanso has joined #openstack-infra | 02:16 | |
| *** ddieterl_ has joined #openstack-infra | 02:17 | |
| greghaynes | TheJulia: also, other note would be to make sure kvm is enabled in the bios of the system youre using ;) | 02:18 |
| *** ddieterly has quit IRC | 02:18 | |
| greghaynes | I might have ran into that on my laptop recently which is why I recognize the error | 02:18 |
| *** tiswanso has quit IRC | 02:19 | |
| *** tiswanso has joined #openstack-infra | 02:20 | |
| TheJulia | greghaynes: new laptop? | 02:21 |
| greghaynes | Yep, X1 v3 | 02:21 |
| greghaynes | and it even has linux drivers now! | 02:21 |
| *** otter768 has joined #openstack-infra | 02:22 | |
| TheJulia | nice | 02:22 |
| tchaypo | TheJulia: might i suggest you try coffee in the morning instead? | 02:23 |
| TheJulia | Yes... That would be an excellent idea! | 02:26 |
| *** otter768 has quit IRC | 02:26 | |
| *** sputnik13 has joined #openstack-infra | 02:27 | |
| *** esker has joined #openstack-infra | 02:29 | |
| *** baoli has joined #openstack-infra | 02:30 | |
| *** koolhead17 has joined #openstack-infra | 02:35 | |
| *** yamahata has quit IRC | 02:39 | |
| *** annegentle has joined #openstack-infra | 02:40 | |
| *** craigbr has joined #openstack-infra | 02:41 | |
| *** annegentle has quit IRC | 02:41 | |
| *** craigbr has quit IRC | 02:42 | |
| *** spzala has quit IRC | 02:42 | |
| *** sarob has quit IRC | 02:42 | |
| clarkb | you cant hvm on our nodes there us no virt passthrough | 02:43 |
| clarkb | you can qemu | 02:43 |
| clarkb | so ya no kvm sorry | 02:43 |
| *** Swami_ has joined #openstack-infra | 02:43 | |
| clarkb | TheJulia greghaynes ^ | 02:44 |
| *** baoli has quit IRC | 02:44 | |
| greghaynes | oh, yea, it didnt occur to me this might be running on a cloud | 02:44 |
| *** weshay has quit IRC | 02:46 | |
| TheJulia | doh! | 02:46 |
| TheJulia | thats what I was very quickly suspecting | 02:47 |
| *** bhunter71 has quit IRC | 02:47 | |
| *** Swami has quit IRC | 02:47 | |
| *** achanda has joined #openstack-infra | 02:47 | |
| *** esker has quit IRC | 02:47 | |
| *** gyee has quit IRC | 02:47 | |
| *** sdake has quit IRC | 02:51 | |
| *** ashleighfarnham has quit IRC | 02:54 | |
| *** unicell has quit IRC | 02:58 | |
| *** alexpilotti has quit IRC | 03:00 | |
| *** signed8b_ is now known as signed8bit_ZZZzz | 03:01 | |
| *** baoli has joined #openstack-infra | 03:02 | |
| *** mahito has quit IRC | 03:02 | |
| *** yamahata has joined #openstack-infra | 03:04 | |
| *** chlong has quit IRC | 03:05 | |
| *** annegentle has joined #openstack-infra | 03:10 | |
| *** aduarte has quit IRC | 03:12 | |
| *** esker has joined #openstack-infra | 03:13 | |
| *** panda has quit IRC | 03:13 | |
| *** panda has joined #openstack-infra | 03:14 | |
| *** aduarte has joined #openstack-infra | 03:15 | |
| *** signed8bit_ZZZzz is now known as signed8b_ | 03:16 | |
| *** salv-orlando has joined #openstack-infra | 03:17 | |
| *** signed8b_ is now known as signed8bit_ZZZzz | 03:17 | |
| *** dims_ has quit IRC | 03:18 | |
| *** annegentle has quit IRC | 03:20 | |
| *** esker has quit IRC | 03:24 | |
| *** yamamoto has quit IRC | 03:24 | |
| *** esker has joined #openstack-infra | 03:24 | |
| *** btully has quit IRC | 03:28 | |
| *** tiswanso has quit IRC | 03:29 | |
| *** links has joined #openstack-infra | 03:30 | |
| *** camunoz has quit IRC | 03:31 | |
| *** esker has quit IRC | 03:32 | |
| *** ddieterl_ has quit IRC | 03:33 | |
| *** dimtruck is now known as zz_dimtruck | 03:36 | |
| *** marun has quit IRC | 03:41 | |
| *** patrickeast has joined #openstack-infra | 03:42 | |
| *** salv-orlando has quit IRC | 03:42 | |
| *** trey has quit IRC | 03:43 | |
| *** shashankhegde has quit IRC | 03:44 | |
| *** trey has joined #openstack-infra | 03:45 | |
| *** camunoz has joined #openstack-infra | 03:46 | |
| openstackgerrit | John Griffith proposed openstack-infra/project-config: Add functional test job for cinderclient https://review.openstack.org/182528 | 03:47 |
| *** david-lyle has joined #openstack-infra | 03:48 | |
| *** asettle has joined #openstack-infra | 03:49 | |
| *** baoli has quit IRC | 03:49 | |
| *** puranamr has joined #openstack-infra | 03:49 | |
| *** koolhead17 has quit IRC | 03:51 | |
| *** chlong_ has quit IRC | 03:52 | |
| *** Sukhdev_ has quit IRC | 03:53 | |
| *** puranamr has quit IRC | 03:55 | |
| *** coolsvap|afk is now known as coolsvap | 03:56 | |
| *** BharatK has joined #openstack-infra | 03:56 | |
| *** banix has quit IRC | 04:03 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 04:06 | |
| *** yamamoto has joined #openstack-infra | 04:07 | |
| *** yfried_ has joined #openstack-infra | 04:08 | |
| *** salv-orlando has joined #openstack-infra | 04:10 | |
| *** yfried has quit IRC | 04:10 | |
| *** sarob has joined #openstack-infra | 04:12 | |
| *** wenlock has joined #openstack-infra | 04:14 | |
| *** yamamoto has quit IRC | 04:15 | |
| *** hdd_ has quit IRC | 04:16 | |
| *** fedexo has quit IRC | 04:17 | |
| *** yfried_ has quit IRC | 04:17 | |
| *** sarob has quit IRC | 04:18 | |
| *** dims has joined #openstack-infra | 04:18 | |
| *** sarob has joined #openstack-infra | 04:19 | |
| *** yamamoto has joined #openstack-infra | 04:22 | |
| *** yamamoto has quit IRC | 04:22 | |
| *** otter768 has joined #openstack-infra | 04:22 | |
| *** ildikov has quit IRC | 04:23 | |
| *** salv-orlando has quit IRC | 04:25 | |
| *** dims has quit IRC | 04:25 | |
| openstackgerrit | Swaminathan Vasudevan proposed openstack-infra/project-config: Make neutron-dvr job voting at the gate https://review.openstack.org/180230 | 04:26 |
| *** otter768 has quit IRC | 04:27 | |
| *** sarob has quit IRC | 04:30 | |
| *** salv-orlando has joined #openstack-infra | 04:32 | |
| *** salv-orlando has quit IRC | 04:33 | |
| *** sarob has joined #openstack-infra | 04:34 | |
| *** ddieterly has joined #openstack-infra | 04:34 | |
| *** ildikov has joined #openstack-infra | 04:36 | |
| *** ddieterly has quit IRC | 04:38 | |
| openstackgerrit | James Polley proposed openstack-dev/pbr: WIP: Teach pbr to read extras from setup.cfg https://review.openstack.org/181795 | 04:41 |
| *** deepakcs has joined #openstack-infra | 04:41 | |
| *** ildikov has quit IRC | 04:43 | |
| *** hichihara has quit IRC | 04:46 | |
| *** dannywilson has joined #openstack-infra | 04:47 | |
| *** dannywilson has quit IRC | 04:49 | |
| *** sarob has quit IRC | 04:49 | |
| *** dannywilson has joined #openstack-infra | 04:49 | |
| *** markvoelker has quit IRC | 04:58 | |
| *** ildikov has joined #openstack-infra | 04:59 | |
| *** CongTo has joined #openstack-infra | 04:59 | |
| *** CongToThanh has joined #openstack-infra | 05:01 | |
| *** CongToThanh has quit IRC | 05:02 | |
| *** patrickeast has quit IRC | 05:03 | |
| *** ir2ivps has quit IRC | 05:04 | |
| *** markvoelker has joined #openstack-infra | 05:06 | |
| *** unicell has joined #openstack-infra | 05:07 | |
| *** Swami__ has joined #openstack-infra | 05:10 | |
| *** unicell has quit IRC | 05:11 | |
| *** unicell has joined #openstack-infra | 05:12 | |
| *** achanda_ has joined #openstack-infra | 05:12 | |
| *** hemna_ has joined #openstack-infra | 05:13 | |
| *** SlickN1k has joined #openstack-infra | 05:14 | |
| *** funzo_ has joined #openstack-infra | 05:14 | |
| *** katyafervent_ has joined #openstack-infra | 05:14 | |
| *** stevemar2 has joined #openstack-infra | 05:14 | |
| *** Swami_ has quit IRC | 05:14 | |
| *** mikap has joined #openstack-infra | 05:15 | |
| *** hughsaunders_ has joined #openstack-infra | 05:16 | |
| *** btully has joined #openstack-infra | 05:16 | |
| *** SergK_ has joined #openstack-infra | 05:16 | |
| lifeless | I presume fungi has halt()ed | 05:16 |
| lifeless | jogo: / clarkb: whats the status on unwedging the gate? | 05:17 |
| lifeless | I see 182438 is still stalled | 05:17 |
| *** rbradfor_ has joined #openstack-infra | 05:17 | |
| *** mikal_ has joined #openstack-infra | 05:17 | |
| *** dosaboy_ has joined #openstack-infra | 05:18 | |
| *** yfried_ has joined #openstack-infra | 05:19 | |
| *** bboese_ has joined #openstack-infra | 05:20 | |
| greghaynes | lifeless: jogo posted https://review.openstack.org/#/c/182526/ thinking it would unwedge the gate before I think he EOD'd, looks like it didnt pass | 05:20 |
| *** barnaby has joined #openstack-infra | 05:21 | |
| greghaynes | lifeless: looks like relevant backscroll is in #openstack-qa | 05:22 |
| *** skraynev_ has joined #openstack-infra | 05:22 | |
| *** ir2ivps has joined #openstack-infra | 05:22 | |
| lifeless | I thought https://review.openstack.org/182524 was the thing he proposed per the list | 05:23 |
| *** coolsvap- has joined #openstack-infra | 05:23 | |
| *** chlong has joined #openstack-infra | 05:23 | |
| *** fawadkhaliq has joined #openstack-infra | 05:24 | |
| greghaynes | lifeless: Ah, yep, that was a few lines up in backscroll | 05:25 |
| *** achanda has quit IRC | 05:26 | |
| *** stevemar has quit IRC | 05:26 | |
| *** rbradfor has quit IRC | 05:26 | |
| *** bcourt has quit IRC | 05:26 | |
| *** coolsvap has quit IRC | 05:26 | |
| *** funzo has quit IRC | 05:26 | |
| *** hemna has quit IRC | 05:26 | |
| *** anteaya has quit IRC | 05:26 | |
| *** bboese has quit IRC | 05:27 | |
| *** SergK has quit IRC | 05:27 | |
| *** SlickNik has quit IRC | 05:27 | |
| *** dosaboy has quit IRC | 05:27 | |
| *** katyafervent has quit IRC | 05:27 | |
| *** hughsaunders has quit IRC | 05:27 | |
| *** mika has quit IRC | 05:27 | |
| *** skraynev has quit IRC | 05:27 | |
| *** mikal has quit IRC | 05:27 | |
| *** SlickN1k is now known as SlickNik | 05:27 | |
| *** hughsaunders_ is now known as hughsaunders | 05:27 | |
| *** rbradfor_ is now known as rbradfor | 05:27 | |
| *** mikap is now known as mika | 05:27 | |
| *** anteaya has joined #openstack-infra | 05:28 | |
| *** CongTo has quit IRC | 05:29 | |
| *** dalgaaf has joined #openstack-infra | 05:30 | |
| *** mrmartin has joined #openstack-infra | 05:30 | |
| *** yamamoto has joined #openstack-infra | 05:33 | |
| *** chlong has quit IRC | 05:34 | |
| *** ddieterly has joined #openstack-infra | 05:35 | |
| *** sks has joined #openstack-infra | 05:37 | |
| *** ddieterly has quit IRC | 05:39 | |
| *** ibiris_away is now known as ibiris | 05:40 | |
| *** ashleighfarnham has joined #openstack-infra | 05:43 | |
| *** claudiub has joined #openstack-infra | 05:45 | |
| *** ildikov has quit IRC | 05:45 | |
| *** chlong has joined #openstack-infra | 05:48 | |
| *** camunoz has quit IRC | 05:48 | |
| *** claudiub has quit IRC | 05:51 | |
| *** rlandy has joined #openstack-infra | 05:52 | |
| *** mrunge has joined #openstack-infra | 05:52 | |
| *** wenlock has quit IRC | 05:55 | |
| *** yfried_ is now known as yfried|afk | 05:58 | |
| *** ashleighfarnham has quit IRC | 05:59 | |
| *** ashleighfarnham has joined #openstack-infra | 05:59 | |
| *** camunoz has joined #openstack-infra | 06:00 | |
| jogo | lifeless: that was a different patch | 06:00 |
| jogo | lifeless: this is the one https://review.openstack.org/#/c/182524/ | 06:00 |
| *** dims has joined #openstack-infra | 06:01 | |
| *** yfried|afk is now known as yfried_ | 06:01 | |
| *** BharatK has quit IRC | 06:01 | |
| *** maishsk has joined #openstack-infra | 06:02 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack-infra/project-config: Normalize projects.yaml https://review.openstack.org/182541 | 06:04 |
| openstackgerrit | Wayne Warren proposed openstack-infra/infra-specs: Jenkins Job Builder 2.0.0 API Changes https://review.openstack.org/182542 | 06:04 |
| *** ashleighfarnham has quit IRC | 06:04 | |
| openstackgerrit | Wayne Warren proposed openstack-infra/infra-specs: Jenkins Job Builder 2.0.0 API Changes https://review.openstack.org/182542 | 06:06 |
| *** dims has quit IRC | 06:06 | |
| *** stevemar2 is now known as stevemar | 06:10 | |
| *** afazekas_ has joined #openstack-infra | 06:11 | |
| *** yfried_ is now known as yfried|afk | 06:12 | |
| *** yfried|afk is now known as yfried_ | 06:12 | |
| *** BharatK has joined #openstack-infra | 06:16 | |
| *** tnovacik has joined #openstack-infra | 06:20 | |
| lifeless | jogo: thats the one I thought it was ;) | 06:22 |
| *** mpaolino has joined #openstack-infra | 06:23 | |
| *** otter768 has joined #openstack-infra | 06:23 | |
| *** chlong has quit IRC | 06:24 | |
| *** jcoufal has joined #openstack-infra | 06:25 | |
| *** chlong has joined #openstack-infra | 06:27 | |
| openstackgerrit | Wayne Warren proposed openstack-infra/infra-specs: Develop gitub-integration spec. https://review.openstack.org/163637 | 06:28 |
| *** otter768 has quit IRC | 06:28 | |
| openstackgerrit | Merged openstack-infra/project-config: Normalize projects.yaml https://review.openstack.org/182541 | 06:32 |
| *** shardy has joined #openstack-infra | 06:32 | |
| *** ildikov has joined #openstack-infra | 06:32 | |
| *** dannywilson has quit IRC | 06:33 | |
| *** ddieterly has joined #openstack-infra | 06:36 | |
| *** asettle has quit IRC | 06:38 | |
| *** soren has joined #openstack-infra | 06:38 | |
| *** ddieterly has quit IRC | 06:40 | |
| *** stevemar has quit IRC | 06:43 | |
| *** BharatK has quit IRC | 06:44 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/puppet-cgit: Updated config for CentOS 7 and Apache 2.4 https://review.openstack.org/182394 | 06:45 |
| *** smallbig has left #openstack-infra | 06:49 | |
| *** skraynev_ is now known as skraynev | 06:54 | |
| *** achanda_ has quit IRC | 06:55 | |
| *** BharatK has joined #openstack-infra | 06:55 | |
| *** HeOS has quit IRC | 06:55 | |
| *** fifieldt has joined #openstack-infra | 06:56 | |
| *** fifieldt has quit IRC | 06:56 | |
| *** scheuran has joined #openstack-infra | 06:56 | |
| openstackgerrit | Yanis Guenane proposed openstack-infra/tripleo-ci: Fix path issue to controller-config-pacemaker https://review.openstack.org/182553 | 07:00 |
| *** woodster_ has quit IRC | 07:00 | |
| *** markus_z has joined #openstack-infra | 07:06 | |
| *** ZZelle_ has quit IRC | 07:08 | |
| *** ZZelle_ has joined #openstack-infra | 07:08 | |
| *** Swami__ has quit IRC | 07:09 | |
| *** ildikov has quit IRC | 07:10 | |
| *** yfried_ is now known as yfried|afk | 07:10 | |
| *** panda has quit IRC | 07:13 | |
| *** panda has joined #openstack-infra | 07:14 | |
| *** mrunge has quit IRC | 07:14 | |
| *** mpavone has joined #openstack-infra | 07:18 | |
| *** dizquierdo has joined #openstack-infra | 07:20 | |
| *** yfried|afk is now known as yfried_ | 07:22 | |
| *** tnovacik has quit IRC | 07:24 | |
| *** ildikov has joined #openstack-infra | 07:25 | |
| *** arxcruz has joined #openstack-infra | 07:28 | |
| *** yamamoto has quit IRC | 07:31 | |
| *** ajmiller__ has quit IRC | 07:33 | |
| *** Hal1 has joined #openstack-infra | 07:34 | |
| *** ajmiller has joined #openstack-infra | 07:34 | |
| *** ddieterly has joined #openstack-infra | 07:37 | |
| *** jlanoux has joined #openstack-infra | 07:38 | |
| *** ddieterly has quit IRC | 07:41 | |
| openstackgerrit | Jaume Devesa proposed openstack-infra/project-config: Make `networking-midonet` tempest test vote https://review.openstack.org/182560 | 07:43 |
| *** jtomasek has joined #openstack-infra | 07:44 | |
| *** jistr has joined #openstack-infra | 07:49 | |
| *** btully has quit IRC | 07:49 | |
| *** dguerri is now known as _dguerri | 07:52 | |
| *** _dguerri is now known as dguerri | 07:52 | |
| *** mmedvede has quit IRC | 07:56 | |
| *** mpavone has quit IRC | 07:57 | |
| *** Longgeek has joined #openstack-infra | 08:00 | |
| *** dizquierdo has quit IRC | 08:02 | |
| *** derekh has joined #openstack-infra | 08:03 | |
| *** mpavone has joined #openstack-infra | 08:04 | |
| *** yamamoto has joined #openstack-infra | 08:06 | |
| *** yamamoto has quit IRC | 08:08 | |
| *** mmedvede has joined #openstack-infra | 08:11 | |
| *** mrunge has joined #openstack-infra | 08:11 | |
| *** sergsh has joined #openstack-infra | 08:13 | |
| *** HeOS has joined #openstack-infra | 08:20 | |
| *** mrmartin has quit IRC | 08:20 | |
| *** otter768 has joined #openstack-infra | 08:24 | |
| *** e0ne has joined #openstack-infra | 08:25 | |
| *** e0ne has quit IRC | 08:28 | |
| *** devvesa has joined #openstack-infra | 08:29 | |
| *** otter768 has quit IRC | 08:29 | |
| *** e0ne has joined #openstack-infra | 08:30 | |
| *** nfedotov has joined #openstack-infra | 08:31 | |
| *** e0ne has quit IRC | 08:31 | |
| *** pal has joined #openstack-infra | 08:33 | |
| *** ddieterly has joined #openstack-infra | 08:37 | |
| *** tnovacik has joined #openstack-infra | 08:38 | |
| *** ddieterly has quit IRC | 08:42 | |
| *** hashar has joined #openstack-infra | 08:42 | |
| *** dguitarbite has joined #openstack-infra | 08:43 | |
| *** yamamoto has joined #openstack-infra | 08:49 | |
| *** yamamoto has quit IRC | 08:51 | |
| *** ssam2 has joined #openstack-infra | 08:59 | |
| *** dizquierdo has joined #openstack-infra | 09:02 | |
| *** che-arne has joined #openstack-infra | 09:07 | |
| lifeless | jd__: the gate is doubly wedged for pbr | 09:10 |
| *** ihrachyshka has joined #openstack-infra | 09:10 | |
| openstackgerrit | Merged openstack-infra/irc-meetings: Remove Blazar Meeting https://review.openstack.org/182509 | 09:11 |
| openstackgerrit | Merged openstack-infra/irc-meetings: Move Cue meeting to Monday https://review.openstack.org/182510 | 09:11 |
| lifeless | jd__: neutron's trunk is un-installable (see my mail to -dev) and pbr's integration tests run that | 09:11 |
| jd__ | lifeless: ah thanks for the heads-up | 09:11 |
| openstackgerrit | Merged openstack-infra/irc-meetings: Update Neutron Service chaining Meeting https://review.openstack.org/182516 | 09:11 |
| lifeless | jd__: and neutron's fix can't land cause grenade is bust (see joe's mail to -dev) | 09:11 |
| jd__ | I've a bit of backlog on -dev | 09:11 |
| lifeless | oh, its been fun :) | 09:11 |
| jd__ | I bet :( | 09:12 |
| lifeless | jd__: OTOH I finally pulled together a blog post about the packaging plumbing stuff | 09:16 |
| lifeless | jd__: which was satisfying | 09:16 |
| *** mrmartin has joined #openstack-infra | 09:17 | |
| jd__ | lifeless: it's on my to-read list yeah I saw that :) | 09:17 |
| *** e0ne has joined #openstack-infra | 09:22 | |
| openstackgerrit | Jaume Devesa proposed openstack-infra/project-config: Make `networking-midonet` tempest test vote https://review.openstack.org/182560 | 09:25 |
| *** fhubik has joined #openstack-infra | 09:30 | |
| AJaeger | lifeless: thanks for that blog post, it's great to get such overviews! | 09:32 |
| lifeless | yw | 09:32 |
| *** dims has joined #openstack-infra | 09:35 | |
| *** fawadkhaliq has quit IRC | 09:36 | |
| *** e0ne is now known as e0ne_ | 09:36 | |
| *** e0ne_ is now known as e0ne | 09:38 | |
| *** ddieterly has joined #openstack-infra | 09:38 | |
| *** yfried_ is now known as yfried|afk | 09:39 | |
| *** dims has quit IRC | 09:40 | |
| *** ddieterly has quit IRC | 09:43 | |
| *** ildikov has quit IRC | 09:48 | |
| *** yfried|afk is now known as yfried_ | 09:59 | |
| *** dims has joined #openstack-infra | 10:01 | |
| *** dizquierdo has quit IRC | 10:01 | |
| openstackgerrit | Julien Danjou proposed openstack-infra/git-review: Allow to specify which commit to send https://review.openstack.org/139701 | 10:01 |
| *** ildikov has joined #openstack-infra | 10:03 | |
| *** mrmartin has quit IRC | 10:06 | |
| *** jamielennox is now known as jamielennox|away | 10:08 | |
| *** pc_m has joined #openstack-infra | 10:09 | |
| *** doude has joined #openstack-infra | 10:09 | |
| *** dizquierdo has joined #openstack-infra | 10:14 | |
| openstackgerrit | Nikki Heald proposed openstack-infra/pypi-mirror: If we specify a branch, always use it https://review.openstack.org/182608 | 10:15 |
| *** fhubik is now known as fhubik_afk | 10:16 | |
| openstackgerrit | Julien Danjou proposed openstack-infra/git-review: Allow to specify which commit to send https://review.openstack.org/139701 | 10:20 |
| *** fhubik_afk is now known as fhubik | 10:22 | |
| *** notnownikki has joined #openstack-infra | 10:23 | |
| *** e0ne is now known as e0ne_ | 10:25 | |
| *** avozza has quit IRC | 10:25 | |
| *** melwitt has quit IRC | 10:25 | |
| *** otter768 has joined #openstack-infra | 10:25 | |
| *** mattoliverau has quit IRC | 10:25 | |
| *** jhesketh has quit IRC | 10:25 | |
| *** wendar has quit IRC | 10:25 | |
| *** gus has quit IRC | 10:25 | |
| *** wendar_ has joined #openstack-infra | 10:25 | |
| *** ianw has quit IRC | 10:26 | |
| *** jbernard has quit IRC | 10:26 | |
| *** e0ne_ is now known as e0ne | 10:26 | |
| *** spredzy has quit IRC | 10:26 | |
| *** mattoliverau has joined #openstack-infra | 10:26 | |
| *** taron has quit IRC | 10:27 | |
| *** tlbr has quit IRC | 10:27 | |
| *** gaelL has quit IRC | 10:27 | |
| *** bradm_ has joined #openstack-infra | 10:27 | |
| *** bradm has quit IRC | 10:27 | |
| *** melwitt has joined #openstack-infra | 10:28 | |
| *** mikal_ has quit IRC | 10:28 | |
| *** mikal has joined #openstack-infra | 10:28 | |
| *** hashar is now known as hasharLunch | 10:28 | |
| *** jd__ has quit IRC | 10:28 | |
| *** zhiwei has quit IRC | 10:29 | |
| *** aarefiev has quit IRC | 10:29 | |
| *** mhu has quit IRC | 10:29 | |
| *** gaelL has joined #openstack-infra | 10:30 | |
| *** aarefiev has joined #openstack-infra | 10:30 | |
| *** otter768 has quit IRC | 10:30 | |
| *** slypushenko_ has quit IRC | 10:30 | |
| *** plol has quit IRC | 10:31 | |
| *** spredzy has joined #openstack-infra | 10:31 | |
| *** plol has joined #openstack-infra | 10:31 | |
| *** mhu has joined #openstack-infra | 10:31 | |
| *** jhesketh has joined #openstack-infra | 10:31 | |
| *** grantbow has quit IRC | 10:31 | |
| *** slypushenko_ has joined #openstack-infra | 10:31 | |
| *** ivar-laz_ has quit IRC | 10:32 | |
| *** jd__ has joined #openstack-infra | 10:32 | |
| *** grantbow has joined #openstack-infra | 10:32 | |
| *** zz_avozza has joined #openstack-infra | 10:32 | |
| *** EmilienM has quit IRC | 10:33 | |
| *** zz_avozza is now known as avozza | 10:33 | |
| *** yamamoto has joined #openstack-infra | 10:33 | |
| *** tlbr has joined #openstack-infra | 10:33 | |
| *** ekarlso has quit IRC | 10:33 | |
| *** gus has joined #openstack-infra | 10:33 | |
| *** aarefiev has quit IRC | 10:35 | |
| *** EmilienM has joined #openstack-infra | 10:35 | |
| *** aarefiev has joined #openstack-infra | 10:35 | |
| *** taron has joined #openstack-infra | 10:36 | |
| *** jbernard has joined #openstack-infra | 10:36 | |
| *** ddieterly has joined #openstack-infra | 10:39 | |
| samueldmq | morning :) | 10:39 |
| *** e0ne is now known as e0ne_ | 10:39 | |
| *** yamamoto has quit IRC | 10:39 | |
| *** e0ne_ is now known as e0ne | 10:40 | |
| *** ekarlso has joined #openstack-infra | 10:40 | |
| *** dizquierdo has quit IRC | 10:41 | |
| *** deepakcs has quit IRC | 10:41 | |
| *** mdenny has quit IRC | 10:41 | |
| *** ddieterly has quit IRC | 10:43 | |
| *** mpaolino has quit IRC | 10:44 | |
| *** ekarlso has quit IRC | 10:45 | |
| *** ekarlso has joined #openstack-infra | 10:45 | |
| *** yamamoto has joined #openstack-infra | 10:46 | |
| *** dizquierdo has joined #openstack-infra | 10:54 | |
| *** yfried_ has quit IRC | 10:55 | |
| *** tmcpeak has joined #openstack-infra | 10:56 | |
| *** kmartin has quit IRC | 10:56 | |
| *** yfried_ has joined #openstack-infra | 10:56 | |
| *** Trozz has quit IRC | 10:58 | |
| *** asselin has quit IRC | 11:01 | |
| *** ildikov has quit IRC | 11:02 | |
| *** Trozz has joined #openstack-infra | 11:05 | |
| *** pal has quit IRC | 11:07 | |
| *** dalgaaf has quit IRC | 11:07 | |
| *** doug-fish has joined #openstack-infra | 11:12 | |
| *** tnovacik has quit IRC | 11:12 | |
| *** ociuhandu has joined #openstack-infra | 11:13 | |
| *** cdent has joined #openstack-infra | 11:13 | |
| *** panda has quit IRC | 11:13 | |
| pelix | jeblair zaro mordred waynr: when you get in, you might take a look at https://review.openstack.org/181364 which is improving the performance of JJB by a whooping margin (order of magnitude improvement) by reducing the number of queries for the list of jobs in Jenkins from per job, to a single query at the start. | 11:14 |
| pelix | Think that should be landed ASAP | 11:14 |
| *** panda has joined #openstack-infra | 11:14 | |
| *** weshay has joined #openstack-infra | 11:15 | |
| pelix | fungi: ^^ think that might allow you to run with the threading patch subsequently without bringing the jenkins master to it's knees | 11:16 |
| *** ildikov has joined #openstack-infra | 11:17 | |
| *** julim has quit IRC | 11:18 | |
| openstackgerrit | Samuel de Medeiros Queiroz proposed openstack-infra/project-config: Experimental Devstack job for Identity v3 API only https://review.openstack.org/179661 | 11:25 |
| *** e0ne is now known as e0ne_ | 11:26 | |
| *** miqui has quit IRC | 11:28 | |
| *** salv-orlando has joined #openstack-infra | 11:28 | |
| *** ivar-lazzaro has joined #openstack-infra | 11:32 | |
| *** ldnunes has joined #openstack-infra | 11:33 | |
| *** mjturek1 has joined #openstack-infra | 11:34 | |
| *** ivar-lazzaro has quit IRC | 11:37 | |
| *** salv-orlando has quit IRC | 11:37 | |
| notnownikki | i'm getting failures on check-requirements-integration-dsvm that seem unrelated to my change, can I get some help debugging please? change is at https://review.openstack.org/#/c/182608/ , failure seems to result from http://logs.openstack.org/08/182608/1/check/check-requirements-integration-dsvm/e11e1bf/console.html#_2015-05-13_11_24_11_673 | 11:37 |
| sdague | zaro: so, I see that the groups fix for gerrit is abandoned - https://gerrit-review.googlesource.com/#/c/62880/1 - is there someway we can get that moving again? I feel like that's a big constrictor in the current review workflow | 11:37 |
| sdague | notnownikki: there is an ML thread from lifeless about pbr vs. neutron | 11:38 |
| sdague | I guess that is not yet resolved | 11:38 |
| notnownikki | oh :( | 11:39 |
| *** fhubik is now known as fhubik_afk | 11:39 | |
| *** psedlak has joined #openstack-infra | 11:39 | |
| *** ddieterly has joined #openstack-infra | 11:40 | |
| *** mrmartin has joined #openstack-infra | 11:40 | |
| *** coolsvap- is now known as coolsvap|afk | 11:41 | |
| *** salv-orlando has joined #openstack-infra | 11:43 | |
| *** ddieterly has quit IRC | 11:45 | |
| *** mwagner_lap has quit IRC | 11:46 | |
| *** yamahata has quit IRC | 11:48 | |
| *** hasharLunch has quit IRC | 11:51 | |
| *** BharatK has quit IRC | 11:53 | |
| *** sdake has joined #openstack-infra | 11:54 | |
| *** dprince has joined #openstack-infra | 11:54 | |
| mordred | sdague: anything I need to do related to the neutron/pbr before I do other things this morning? | 11:56 |
| mordred | lifeless: ^^? | 11:56 |
| fungi | wondering if the neutron patch got rechecked after the grenade fix (182526) landed | 11:58 |
| sdague | mordred: I do not know | 11:58 |
| *** woodster_ has joined #openstack-infra | 11:59 | |
| *** sdake has quit IRC | 11:59 | |
| sdague | I'm actually not very up on the current break status, that mostly started failing at my end of day | 11:59 |
| *** sdake has joined #openstack-infra | 12:00 | |
| fungi | same. i was starting to fade when it all started up | 12:01 |
| mordred | sdague: same here | 12:01 |
| mordred | awesome | 12:01 |
| fungi | 182438 made it into the gate 20 minutes ago | 12:01 |
| mordred | well, it's good that the three of us grok it all :) | 12:01 |
| *** Krinkle|detached is now known as Krinkle | 12:01 | |
| fungi | unless the trove change ahead of it fails, should merge in ~13 minutes | 12:02 |
| notnownikki | yay :D | 12:02 |
| mordred | pelix: done. WOW | 12:03 |
| mordred | fungi: you should read pelix's final review comment on 181364 btw | 12:04 |
| samueldmq | morning, any core on d-g who could approve 'Remove arbitrary devstack config options' ? https://review.openstack.org/#/c/179982/ | 12:05 |
| *** BharatK has joined #openstack-infra | 12:05 | |
| samueldmq | it already has a +2 from mordred, but I guess it needs another one | 12:05 |
| samueldmq | the project-config change on that already merged days ago | 12:05 |
| fungi | wow--job queries invalidate the jenkins cache? craaaaazy | 12:05 |
| openstackgerrit | Merged openstack-infra/jenkins-job-builder: Query all jobs once https://review.openstack.org/181364 | 12:06 |
| mordred | yolanda: ^^ nice work | 12:06 |
| *** e0ne_ is now known as e0ne | 12:08 | |
| openstackgerrit | Samuel de Medeiros Queiroz proposed openstack-infra/project-config: Experimental Devstack job for Identity v3 API only https://review.openstack.org/179661 | 12:08 |
| *** derekh_ has joined #openstack-infra | 12:08 | |
| mordred | fungi: btw - clarkb finished testing 168603 - so I think we can land it now | 12:10 |
| samueldmq | mordred, yes that is nice .. but in https://review.openstack.org/#/c/181364/7/jenkins_jobs/builder.py | 12:11 |
| samueldmq | mordred, L 173, wouldn't it be useful if we invalidate the cache and refreshed it ? | 12:11 |
| samueldmq | :-) | 12:11 |
| yolanda | hi mordred, thx | 12:12 |
| *** derekh has quit IRC | 12:12 | |
| yolanda | super fast approval on this one :) | 12:12 |
| mordred | samueldmq: yes - it would. maybe yolanda will make another patch to do that | 12:13 |
| yolanda | mm, let me check | 12:14 |
| *** signed8bit_ZZZzz has quit IRC | 12:14 | |
| *** bswartz has quit IRC | 12:14 | |
| samueldmq | mordred, yolanda nice | 12:15 |
| *** trey has quit IRC | 12:15 | |
| *** fhubik_afk is now known as fhubik | 12:15 | |
| *** dizquierdo has quit IRC | 12:16 | |
| *** trey has joined #openstack-infra | 12:16 | |
| yolanda | samueldmq, wel, it was trying to use jenkins as a fallback there | 12:16 |
| yolanda | if for some reason cache is empty or fails to return for some problem, still go to jenkins and chck | 12:16 |
| yolanda | it shouldn't happen, but just in case | 12:17 |
| samueldmq | yolanda, yeah but in that case, we've detected a cache inconsistency, right ? | 12:17 |
| mordred | yolanda: yah. I think he's saying that if you get a result back from jenkins, maybe you should update the local cache? | 12:17 |
| samueldmq | yolanda, if so, fix up the cache | 12:17 |
| yolanda | samueldmq, yes, in that case is due to cache inconsistency | 12:17 |
| yolanda | it will be worth a retry there, yes | 12:17 |
| samueldmq | mordred, exactly | 12:17 |
| yolanda | if that fails for timeout or some temporary jenkins problem | 12:18 |
| mordred | fungi, sdague: also - while I'm bothering you - 182061 would help me out if I could get it in today | 12:18 |
| yolanda | cool, i'll add a patch for that | 12:18 |
| samueldmq | yolanda, ++ thanks | 12:18 |
| *** yfried__ has joined #openstack-infra | 12:19 | |
| *** psedlak has quit IRC | 12:20 | |
| sdague | mordred: sure | 12:22 |
| sdague | so... the problem with the breaks yesterday, is based on what I saw, if I looked into them in depth I'd mostly want to light people on fire. So I'm staying away. | 12:22 |
| mordred | sdague: yes | 12:22 |
| mordred | sdague: I think lifeless has been doing an excllent job of diving in to them | 12:22 |
| mordred | sdague: and not lighting anyone on fire | 12:23 |
| mordred | so I vote that he continue in his excellent work | 12:23 |
| sdague | yes, that makes me happy :) | 12:23 |
| *** yfried_ has quit IRC | 12:23 | |
| *** mwagner_lap has joined #openstack-infra | 12:25 | |
| *** tiswanso has joined #openstack-infra | 12:25 | |
| *** gordc has joined #openstack-infra | 12:25 | |
| *** isviridov_away is now known as isviridov | 12:25 | |
| *** otter768 has joined #openstack-infra | 12:26 | |
| *** dizquierdo has joined #openstack-infra | 12:28 | |
| *** tiswanso has quit IRC | 12:28 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: Fix ACL(s) for oslo.cache https://review.openstack.org/182645 | 12:29 |
| *** sputnik13 has quit IRC | 12:29 | |
| *** otter768 has quit IRC | 12:31 | |
| *** maishsk has quit IRC | 12:31 | |
| *** alexpilotti has joined #openstack-infra | 12:32 | |
| *** ivar-lazzaro has joined #openstack-infra | 12:33 | |
| *** aysyd has joined #openstack-infra | 12:33 | |
| *** annegentle has joined #openstack-infra | 12:34 | |
| *** mpaolino has joined #openstack-infra | 12:34 | |
| *** prad has joined #openstack-infra | 12:39 | |
| *** salv-orlando has quit IRC | 12:39 | |
| *** bknudson has quit IRC | 12:40 | |
| *** ddieterly has joined #openstack-infra | 12:40 | |
| openstackgerrit | Merged openstack-infra/shade: Use fakes instead of mocks for data objects https://review.openstack.org/182126 | 12:41 |
| *** ivar-lazzaro has quit IRC | 12:41 | |
| ryanpetrello | anybody know why https://review.openstack.org/#/c/159299/ doesn't seem to want to run gate tests and merge? | 12:41 |
| ryanpetrello | it was approved yesterday and never merged | 12:42 |
| ryanpetrello | mordred fungi or sdague ^? | 12:43 |
| *** ddieterly has quit IRC | 12:43 | |
| *** mestery has joined #openstack-infra | 12:44 | |
| *** BharatK has quit IRC | 12:44 | |
| *** annegentle has quit IRC | 12:45 | |
| *** annegentle has joined #openstack-infra | 12:46 | |
| ryanpetrello | weird, I did a -Workflow and then Workflow+1 and it started | 12:47 |
| *** Krinkle is now known as Krinkle|detached | 12:49 | |
| *** kgiusti has joined #openstack-infra | 12:51 | |
| *** dizquierdo has quit IRC | 12:53 | |
| *** julim has joined #openstack-infra | 12:53 | |
| *** hashar has joined #openstack-infra | 12:54 | |
| *** hashar is now known as hasharReview | 12:54 | |
| *** miqui has joined #openstack-infra | 12:54 | |
| *** sambetts has quit IRC | 12:54 | |
| openstackgerrit | Kai Qiang Wu(Kennan) proposed openstack/requirements: Fix the empty string issue for pip https://review.openstack.org/182655 | 12:54 |
| *** electrical has joined #openstack-infra | 12:55 | |
| *** sambetts has joined #openstack-infra | 12:55 | |
| electrical | Hi all, I'm having some issues with the jenkins-job python app. giving a TypeError when its trying to fetch the plugins. running on latest Jenkins with python 3.4.3 | 12:56 |
| *** dizquierdo has joined #openstack-infra | 12:56 | |
| *** banix has joined #openstack-infra | 12:57 | |
| *** banix has quit IRC | 12:58 | |
| electrical | More info about the issue i'm having at https://storyboard.openstack.org/#!/story/2000261 ( sorry for the duplicates in there, storyboard didn't show them up for some reason ) | 12:59 |
| *** sks has quit IRC | 12:59 | |
| *** rkukura has quit IRC | 13:00 | |
| *** rkukura has joined #openstack-infra | 13:01 | |
| fungi | ryanpetrello: not weird. that's the workaround for changes which got approved at some point when zuul wasn't listening to gerrit's event stream or an approval event gets lost. it also triggers when a jenkins +1 vote is applied to a change, but if there was already a +1 on it then gerrit doesn't emit a new vote for replacing an existing vote with one of equal value | 13:01 |
| ryanpetrello | okay, good to know | 13:01 |
| fungi | which is why rechecking those doesn't help | 13:02 |
| *** dkranz has joined #openstack-infra | 13:02 | |
| *** teran has joined #openstack-infra | 13:02 | |
| *** banix has joined #openstack-infra | 13:03 | |
| samueldmq | fungi, does that occur very often ? | 13:04 |
| *** mrmartin has quit IRC | 13:04 | |
| *** banix has quit IRC | 13:04 | |
| samueldmq | fungi, if so we could add something to check the change status (workflow +1 ? ) and then ensure gate jobs have been started | 13:04 |
| fungi | samueldmq: not usually, except for the past couple months we've had a rash of problems with gerrit's event stream hanging and some emergency restarts | 13:04 |
| *** Krinkle|detached is now known as Krinkle | 13:04 | |
| *** dizquierdo has quit IRC | 13:05 | |
| electrical | pelix: zaro yolanda with regards to jenkins-job-builder. perhaps worth to put a note it only works with python 2.7,x? i tried 3.3.x and 3.4.x and doesn't work with all kinds of weird errors :-) | 13:05 |
| *** baoli has joined #openstack-infra | 13:05 | |
| *** baoli has quit IRC | 13:05 | |
| fungi | samueldmq: well, it's more complex than that, because you also need to check a number of other factors to confirm that the change is ready to merge (git dependencies, cross-repo dependencies, et cetera) | 13:05 |
| *** linuxmaniac has joined #openstack-infra | 13:05 | |
| pelix | electrical: I believe there is a bug with using JJB on python 3.4 not related to this change | 13:05 |
| electrical | pelix: ah okay. with 3.3.x it doesn;t work either. with a different error. TypeError: can't use a string pattern on a bytes-like object | 13:06 |
| samueldmq | fungi, well ... not sending it to be merged already ... just ensuring the gate jobs (tests etc) have been started .. wasnt't that the case ? | 13:06 |
| electrical | fails on the same part though | 13:06 |
| pelix | electrical: can you try https://review.openstack.org/#/c/172238/ and see if it fixes that? | 13:06 |
| electrical | pelix: checking it out now. | 13:06 |
| *** annegentle has quit IRC | 13:06 | |
| *** nikil22 has joined #openstack-infra | 13:06 | |
| samueldmq | fungi, whatever need to be done, but avoiding people to do that workaround above ^ | 13:06 |
| fungi | electrical: pelix: that suggests our current python 3.4 test coverage is incomplete, i suppose | 13:06 |
| pelix | electrical: I think the patch might be better being reworked to always use unicode | 13:07 |
| samueldmq | fungi, but as this doesn't happen very often, not sure it's worth it to do | 13:07 |
| *** annegentle has joined #openstack-infra | 13:07 | |
| electrical | fungi: 3.4 and 3.3 | 13:07 |
| fungi | samueldmq: yes, fixing software to avoid outages will avoid that needing to be done. we definitely prefer to spend time eliminating bugs rather than implementing piles of hacky workarounds | 13:07 |
| fungi | electrical: we don't test it on 3.3, only 3.4 | 13:07 |
| electrical | ah okay | 13:07 |
| electrical | i could try out some other versions of 3.x as well if needed | 13:07 |
| *** baoli_ has joined #openstack-infra | 13:07 | |
| pelix | fungi: we're finding out the hard way our coverage is incomplete in a few areas, that change add some tests, and there are others added to check it the CLI with --delete-old in the future that caused the previous regression when the threading patch landed. | 13:08 |
| fungi | pelix: yep, quite familiar with that one ;) | 13:08 |
| electrical | pelix: i dont think that patch will solve it since it fails with fetching data from my jenkins server ( plugin list ) | 13:08 |
| *** ddieterly has joined #openstack-infra | 13:09 | |
| pelix | hmm, wonder if that means there is a bug in python-jenkins with python 3 | 13:09 |
| electrical | pelix: that's my guess at the moment | 13:09 |
| electrical | Let me try out 3.1 as well | 13:10 |
| *** claudiub has joined #openstack-infra | 13:10 | |
| pelix | I suspect unicode issue, so it probably is broken for python >= 3 | 13:10 |
| *** mriedem has quit IRC | 13:10 | |
| fungi | i would look at the traceback in that story, but at the moment storyboard won't let me log in... {"error_description":"You did not provide a valid redirect_uri.","error":"invalid_request"} | 13:11 |
| *** mriedem has joined #openstack-infra | 13:11 | |
| * fungi sighs | 13:11 | |
| fungi | checking logs now | 13:11 |
| electrical | fungi: https://gist.github.com/electrical/0d1c450f805ebc77acb3 | 13:11 |
| *** peristeri has joined #openstack-infra | 13:11 | |
| openstackgerrit | Merged openstack-infra/devstack-gate: Remove arbitrary devstack config options https://review.openstack.org/179982 | 13:11 |
| *** mrmartin has joined #openstack-infra | 13:12 | |
| fungi | [Tue May 12 13:31:06 2015] [error] [client 192.176.1.84] IOError: failed to write data, referer: https://storyboard.openstack.org/ | 13:12 |
| fungi | that's no good | 13:12 |
| electrical | o.0 | 13:12 |
| fungi | oh, wait, that was yesterday | 13:13 |
| electrical | lol :-) | 13:13 |
| *** sks has joined #openstack-infra | 13:13 | |
| fungi | yeah, fungi need coffee | 13:13 |
| *** baoli has joined #openstack-infra | 13:14 | |
| electrical | fungi: pelix TypeError: can't use a string pattern on a bytes-like object on all 3.x Python versions | 13:15 |
| electrical | fails on the get_plugins_info function from the jenkins-python library | 13:15 |
| fungi | yeah, agreed that looks like a python-jenkins bug | 13:16 |
| pelix | electrical: find the line 'response = urlopen(req, timeout=self.timeout).read()' in 'def jenkins_open(...' in the python-jenkins library and change the line to read: | 13:16 |
| pelix | response = urlopen(req, timeout=self.timeout).read().decode("utf-8") | 13:16 |
| *** baoli_ has quit IRC | 13:16 | |
| electrical | pelix: now giving TypeError: must be str, not bytes | 13:17 |
| pelix | I think that may fix it, but obviously means the unit tests for python-jenkins are assuming the type of the response incorrectly on python 3. | 13:18 |
| electrical | but now on the update job | 13:18 |
| electrical | File "/home/richard/.pyenv/versions/3.1.4/lib/python3.1/site-packages/jenkins_jobs/builder.py", line 346, in update_job | 13:18 |
| fungi | apparently we do also gate python-jenkins on python 3.4 unit tests. but that code path must not be traversed (wondering ig we have any coverage over the __init__.py there | 13:18 |
| pelix | fungi: I think the code is covered, but the response type is different between python 2.7 and python 3 which is not accounted for in the tests. | 13:19 |
| electrical | pelix: the error i get now might be solved with the patch you linked me earlier. | 13:19 |
| *** radez_g0n3 is now known as radez | 13:19 | |
| electrical | so i think patching the jekins-python with the utf8 thing and the patch you linked me it should work on python 3.x | 13:19 |
| electrical | i think :-) | 13:19 |
| *** matrohon has joined #openstack-infra | 13:20 | |
| pelix | just need to work out how to update the tests for python-jenkins to show the problem in the first place... | 13:20 |
| electrical | good question | 13:20 |
| fungi | yeah, it seems to just be using the right type in the tests, so misses this class of failure | 13:21 |
| *** mase_x200 has joined #openstack-infra | 13:22 | |
| electrical | I'll use python 2.7.x for now. using pyenv anyway :-) | 13:22 |
| fungi | electrical: fwiw, we're running it under 2.7 in production | 13:23 |
| electrical | I do have an other question while you guys are here. Im hoping to implement this for our environment at Elastic ( Elasticsearch ) we are using some things that there are no definitions for yet. I did read the documentation so i'm going to try and build them. but me being a ruby programmer it might be a slow process :-) | 13:23 |
| *** BharatK has joined #openstack-infra | 13:23 | |
| matrohon | hi all; | 13:24 |
| electrical | especially all the xml stuff is a bit confusing at times for me :-) | 13:24 |
| *** geguileo has joined #openstack-infra | 13:24 | |
| *** tiswanso has joined #openstack-infra | 13:25 | |
| matrohon | I'd like to schedule on tuesday a design session with folks to finalize a spec... | 13:25 |
| matrohon | I'm looking for a suitable location at vancouver for that... | 13:26 |
| matrohon | does anyone can help me? | 13:26 |
| *** Hal1 has quit IRC | 13:26 | |
| fungi | matrohon: i think all of the projects have already selected the topics for their design summit sessions | 13:27 |
| fungi | matrohon: which project is the spec for? you should probably approach the corresponding ptl to discuss further | 13:28 |
| *** hashar has joined #openstack-infra | 13:28 | |
| electrical | fungi: pelix feel free to ping me regarding testing with 3.x I'm gonna have some further fun with the job builder. Hope i can replace all the manual job builds with it :-) | 13:28 |
| *** hasharReview has quit IRC | 13:29 | |
| matrohon | fungi : yep, and we have a slot on friday in neutron contributor meetup, but we would like to have a first meeting earlier, unscheduled | 13:29 |
| pelix | electrical: I think there is a patch to support raw xml directly from yaml blocks where we've yet to add support for something. Not sure it's landed though. | 13:29 |
| fungi | matrohon: the infrastructure team doesn't manage the design summit scheduling | 13:30 |
| electrical | pelix: ah interesting. I rather do it directly the right way so i'll start hacking on it :-) should be a nice learning curve starting from 0 regarding python :-) | 13:30 |
| matrohon | fungi : ok, I was hoping ttx was around to halp me... | 13:30 |
| pelix | electrical: it's only meant as a crutch to keep people moving forward until support is added, just better than being completely blocked. | 13:31 |
| ttx | matrohon: o/ | 13:31 |
| electrical | pelix: very true. in my case its more of a challenge and see if i can convince people internally to move to it. I'm tired of mass changing jobs manually :-) | 13:32 |
| *** tmcpeak has left #openstack-infra | 13:32 | |
| fungi | electrical: that was basically why it got written. openstack had a lot of jobs and the jenkins admins were tired of having to add/modify them or grant random people access to add new ones willy-nilly | 13:33 |
| electrical | fungi: hehe yeah. in some of our cases we have like 100 of the same type of jobs. just different name and github repo. | 13:33 |
| fungi | when we migrated our existing job definitions i want to say we had around 500 and it took something like 3-6 months | 13:34 |
| electrical | wow that's allot | 13:34 |
| fungi | now we have ~5500 jobs | 13:34 |
| electrical | lol :-) | 13:34 |
| fungi | so, it definitely allowed us to scale | 13:34 |
| electrical | hehe yeah indeed | 13:34 |
| *** sputnik13 has joined #openstack-infra | 13:34 | |
| electrical | We currently have 3 different jenkins masters for different projects. would be nice to have a central place to manage it all in an easy way | 13:34 |
| electrical | for example some of my own jobs i have to add different puppet versions to.. not in the mood to change it in every job | 13:35 |
| fungi | definitely. we have 8 jenkins masters (not counting a dev instance to test new jenkins releases and plugins), but they all share a common job set | 13:35 |
| electrical | very handy ::-) | 13:35 |
| electrical | i'm hoping we will be moving to use this | 13:35 |
| fungi | and then we evenly distribute jobs between them via gearman | 13:35 |
| electrical | ahh i see, very nice :-) | 13:36 |
| *** Hal1 has joined #openstack-infra | 13:36 | |
| fungi | because jenkins as we've seen starts to fall over when you go beyond about 100 slaves registered in a given master | 13:36 |
| electrical | wow. we don't have that many yet :p | 13:36 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:37 | |
| fungi | we create and delete slaves on demand, spiking up around 800 at peak usage | 13:37 |
| *** ivar-lazzaro has joined #openstack-infra | 13:37 | |
| electrical | wow okay. that's allot indeed | 13:37 |
| *** claudiub|2 has joined #openstack-infra | 13:38 | |
| electrical | Well, i hope to send some PR's your way soon for adding some plugins to it for the things we use. | 13:38 |
| fungi | at this particular moment we've got about 250 in use running jobs | 13:38 |
| electrical | github PR builder and logstash logging are just 2 of them | 13:38 |
| fungi | but that count will rise as people further west begin to wake up | 13:38 |
| electrical | Ahh, i see | 13:38 |
| matrohon | ttx, fungi : ttx helped me out, thanks | 13:38 |
| *** sdake_ has joined #openstack-infra | 13:38 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add floating IP pool resource methods https://review.openstack.org/180255 | 13:39 |
| fungi | electrical: also, i'm excited to hear that the elasticsearch community may use some of our software! we definitely use some of yours ;) | 13:40 |
| openstackgerrit | Steven Hardy proposed openstack/requirements: Increase stable/kilo python-heatclient upper cap to 0.6 https://review.openstack.org/182672 | 13:40 |
| electrical | fungi: hehe, very cool to hear! | 13:40 |
| *** claudiub has quit IRC | 13:40 | |
| linuxmaniac | zaro: do you still have concerns about https://review.openstack.org/#/c/157735/ ? | 13:40 |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Add glance_store to projects.txt https://review.openstack.org/182674 | 13:40 |
| mriedem | flaper87: ^ | 13:40 |
| electrical | fungi: i do remember something about the CI builds or something that is being logged into ES ? | 13:40 |
| *** mattfarina has joined #openstack-infra | 13:40 | |
| electrical | or was it something else? | 13:40 |
| fungi | electrical: specifically, we have a fairly large elasticsearch cluster backing a logstash instance where we dump the logs form our ci jobs for analysis | 13:41 |
| * flaper87 clicks | 13:41 | |
| *** claudiub has joined #openstack-infra | 13:41 | |
| electrical | fungi: ah, we are planning to do the same. using the logstash reporter in jenkins. or are you using something else? | 13:41 |
| flaper87 | mriedem: thnx | 13:41 |
| fungi | electrical: http://docs.openstack.org/infra/system-config/logstash.html | 13:42 |
| *** ivar-lazzaro has quit IRC | 13:42 | |
| *** claudiub|3 has joined #openstack-infra | 13:42 | |
| *** sdake has quit IRC | 13:42 | |
| *** claudiub|2 has quit IRC | 13:42 | |
| fungi | electrical: homegrown scripts/services to collect, process and inject the logs | 13:42 |
| *** CongTo has joined #openstack-infra | 13:43 | |
| electrical | fungi: ahh i see, okay | 13:43 |
| *** claudiub|2 has joined #openstack-infra | 13:43 | |
| *** yfried__ is now known as yfried|afk | 13:43 | |
| *** claudiub_ has joined #openstack-infra | 13:43 | |
| fungi | electrical: also a service that checks job failures for signs of known bugs based on lucene queries against that data and reports into our code review system on any hits http://docs.openstack.org/infra/system-config/elastic-recheck.html | 13:43 |
| electrical | I've been using the logstash plugin for jenkins. reports quite a nice amount of data including junit stuff | 13:43 |
| yolanda | clarkb, jeblair, i'm dealing with https://review.openstack.org/#/c/76057/ for a downstream sync, and tests are not passing | 13:43 |
| yolanda | i see you both had some comments there | 13:44 |
| *** yfried|afk is now known as yfried__ | 13:44 | |
| yolanda | it's failing for me in all runs, so there is something wrong in this patch | 13:44 |
| mriedem | dhellmann: fungi: sdague: should pbr be the first requirement in requirements.txt for a project? i'm syncing glance_store's requirements with master g-r and it added the comment about how order is important, but i thought that was specifically for pbr being first. | 13:44 |
| *** mrmartin has quit IRC | 13:44 | |
| mriedem | glance_store doesn't have pbr in requirements.txt today | 13:44 |
| electrical | fungi: that's damn handy :-) | 13:44 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add port resource methods https://review.openstack.org/179968 | 13:44 |
| fungi | mriedem: i believe it should... our cookiecutter requirements list it first | 13:45 |
| mriedem | fungi: ok, will add pbr then | 13:45 |
| fungi | mriedem: oh, it only needs it if it imports it somewhere | 13:45 |
| *** claudiub has quit IRC | 13:45 | |
| mriedem | well, it's in setup.py | 13:45 |
| *** dizquierdo has joined #openstack-infra | 13:45 | |
| electrical | fungi: https://www.elastic.co/blog/openstack-elastic-recheck-powered-elk-stack <-- this is why i remembered it :p | 13:46 |
| fungi | mriedem: not having it in the requirements list is not necessarily a bug. having it in setup_requires in setup.py doesn't necessitate adding it to requirements afaik | 13:46 |
| *** claudiub|3 has quit IRC | 13:46 | |
| fungi | electrical: yep, clarkb set up most of that. he'll be awake fairly soon i expect | 13:46 |
| notnownikki | could i get a review on https://review.openstack.org/#/c/182608/ please? it's a really small change that lets us run pypi-mirror against proposed branches of openstack/requirements | 13:46 |
| *** zz_dimtruck is now known as dimtruck | 13:47 | |
| electrical | fungi: hehe cool. | 13:47 |
| mriedem | fungi: ah, ok, i see nova/glance import pbr for version stuff, glance_store doesn't have that so i won't add, thanks | 13:47 |
| *** claudiub|2 has quit IRC | 13:48 | |
| electrical | fungi: pelix thanks for the help so far with jenkins-job generator. will try to send some patches soon-ish for missing plugins :-) | 13:48 |
| *** links has quit IRC | 13:48 | |
| *** hdd has joined #openstack-infra | 13:49 | |
| Shrews | fungi: is it possible we lost some gerrit voting history since yesterday? | 13:50 |
| sdague | mriedem: order is always important, so don't change order if it's working | 13:51 |
| openstackgerrit | Anthony PERARD proposed openstack-infra/devstack-gate: Copy Xen log files after tempest run https://review.openstack.org/182680 | 13:51 |
| mriedem | sdague: i just ran the python update.py from the requirements repo https://review.openstack.org/#/c/182677/1 | 13:51 |
| mriedem | it didn't change order, just added the comment and made me think about pbr | 13:52 |
| *** dboik has joined #openstack-infra | 13:52 | |
| *** dkranz has quit IRC | 13:52 | |
| *** dboik_ has joined #openstack-infra | 13:53 | |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Add glance_store to projects.txt https://review.openstack.org/182681 | 13:53 |
| sdague | mriedem: yeh, so I think there was a version of update.py that stripped comments | 13:54 |
| *** sks has quit IRC | 13:54 | |
| sdague | that caused other issues | 13:54 |
| *** yfried__ is now known as yfried|afk | 13:55 | |
| electrical | fungi: btw, did you guys ever consider to use our official ES puppet module? ;-) | 13:55 |
| *** tnovacik has joined #openstack-infra | 13:56 | |
| *** claudiub_ has quit IRC | 13:57 | |
| *** dboik has quit IRC | 13:57 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone service resource methods https://review.openstack.org/177621 | 13:57 |
| anteaya | Shrews: do you have a patch in question? | 13:57 |
| Shrews | anteaya: https://review.openstack.org/181513 | 13:58 |
| anteaya | Shrews: to the best of my knowledge all gerrit did yesterday is run | 13:58 |
| Shrews | anteaya: i know SpamapS left a +1 on it (still have the email) shortly after uploading patchset 5, yet it is mysteriously missing | 13:58 |
| *** markvoelker has quit IRC | 13:59 | |
| anteaya | Shrews: I see voting history in the comments and current voting table has cleared due to the upload of patchset 5 | 13:59 |
| anteaya | ah | 13:59 |
| anteaya | Shrews: have you a timestamp from the email for when that missing +1 took place? | 13:59 |
| yolanda | clarkb, also increasing timeout to 60 fixed tests for me | 13:59 |
| Shrews | anteaya: 3:15pm EST | 14:00 |
| anteaya | which would have been 19:15 utc if I calculate correctly | 14:00 |
| *** yfried|afk is now known as yfried__ | 14:01 | |
| *** sdake has joined #openstack-infra | 14:01 | |
| anteaya | we were in our infra meeting, now after infra meeting clarkb changed the gerrit db so that stevebaker could use his webui | 14:01 |
| waynr | pelix: nice work! | 14:01 |
| *** mmedvede_ has quit IRC | 14:01 | |
| anteaya | to the best of my knowledge that was the only row changed in teh db | 14:01 |
| anteaya | so here is the log from 19:15 utc yesterday: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2015-05-12.log.html#t2015-05-12T19:15:11 | 14:02 |
| *** Kennan has quit IRC | 14:03 | |
| *** Kennan has joined #openstack-infra | 14:03 | |
| Shrews | anteaya: I wouldn't worry about it too much. I can always ask for a re-vote. If it happens again, though, I'll get more worried. | 14:04 |
| *** sdake_ has quit IRC | 14:04 | |
| anteaya | Shrews: thanks for the alert though, we do need to know if data is getting dropped | 14:06 |
| *** amotoki_ has joined #openstack-infra | 14:06 | |
| anteaya | looking at the logs stevebaker's account was the only one returned in clarkb's query so I'm going with it was the only one that had the date format changed | 14:06 |
| mordred | clarkb, fungi: also: https://review.openstack.org/#/c/182488/ - that came out from yesterday's nodepool testing - I'd like to get it landed before we land the nodepool change | 14:06 |
| *** sks has joined #openstack-infra | 14:07 | |
| *** mtanino has joined #openstack-infra | 14:07 | |
| mmedvede | Anyone knows why https://review.openstack.org/#/c/172007/ still did not merge? Related change https://review.openstack.org/#/c/169956/ is in and now zuul status is broken | 14:07 |
| *** dkranz has joined #openstack-infra | 14:07 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone service resource methods https://review.openstack.org/177621 | 14:10 |
| *** craigbr has joined #openstack-infra | 14:10 | |
| *** claudiub has joined #openstack-infra | 14:11 | |
| *** amitgandhinz has joined #openstack-infra | 14:11 | |
| *** isviridov is now known as isviridov_away | 14:12 | |
| openstackgerrit | Ruby Loo proposed openstack/requirements: Avoid httpretty 0.8.9 as it breaks unittests https://review.openstack.org/182340 | 14:13 |
| *** nikil22 has quit IRC | 14:13 | |
| *** eharney has quit IRC | 14:14 | |
| *** CongTo has quit IRC | 14:14 | |
| *** bknudson has joined #openstack-infra | 14:15 | |
| *** jaypipes has joined #openstack-infra | 14:15 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone endpoint resource methods https://review.openstack.org/180328 | 14:16 |
| *** bswartz has joined #openstack-infra | 14:17 | |
| fungi | mmedvede: i think that approval hit when we were having some gerrit issues. i've reapproved it now | 14:17 |
| fungi | Shrews: what change are you looking at where you're suspecting gerrit voting history issues? | 14:18 |
| mmedvede | fungi: ok, makes sense. Thank you | 14:18 |
| anteaya | fungi: https://review.openstack.org/#/c/181513/ | 14:19 |
| *** craigbr has quit IRC | 14:19 | |
| Shrews | fungi: what anteaya said | 14:19 |
| fungi | electrical: i can't remember if your puppet module existed when clarkb first set this up, but presumably he recalls | 14:19 |
| *** claudiub|2 has joined #openstack-infra | 14:19 | |
| fungi | Shrews: yeah, caught up now | 14:19 |
| electrical | fungi: very possible it didn't exist or was of poor quality :-) | 14:19 |
| electrical | but its been out for almost 2 years | 14:20 |
| electrical | i think :-) | 14:20 |
| *** claudiub|3 has joined #openstack-infra | 14:20 | |
| electrical | 2 Dec 2012 first commit :p | 14:20 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add port resource methods https://review.openstack.org/179968 | 14:21 |
| *** erikmwil_ has joined #openstack-infra | 14:21 | |
| electrical | I hope to be doing a 1.0 release soon-ish | 14:21 |
| *** erikmwil_ is now known as erikmwil | 14:21 | |
| *** claudiub|4 has joined #openstack-infra | 14:21 | |
| *** erikmwil is now known as erikmwilson_ | 14:21 | |
| *** erikmwilson is now known as Guest24723 | 14:21 | |
| *** erikmwilson_ is now known as erikmwilson | 14:21 | |
| *** tonytan4ever has joined #openstack-infra | 14:21 | |
| *** mpaolino has quit IRC | 14:22 | |
| *** erikmwilson_ has joined #openstack-infra | 14:22 | |
| *** claudiub has quit IRC | 14:22 | |
| *** claudiub has joined #openstack-infra | 14:22 | |
| fungi | electrical: yep, definitely predates ours by a couple months: http://git.openstack.org/cgit/openstack-infra/puppet-elasticsearch/log/ (2013-01-29) | 14:23 |
| *** claudiub|2 has quit IRC | 14:23 | |
| electrical | hehe yeah. | 14:23 |
| electrical | Well, just something to think about perhaps worth switching. | 14:24 |
| fungi | well, almost 2 months | 14:24 |
| fungi | electrical: yep, definitely! | 14:24 |
| *** dkranz has quit IRC | 14:24 | |
| electrical | Adding Debian 8 support atm and package pinning | 14:24 |
| fungi | we do like replacing our puppet modules when we can. getting rid of some of the >50 we maintain would be swell | 14:24 |
| electrical | hehe yeah :-) | 14:24 |
| openstackgerrit | Derek Higgins proposed openstack-infra/tripleo-ci: Remove nova temprevert https://review.openstack.org/182695 | 14:25 |
| electrical | Eventually our LS puppet module gets a facelift as well | 14:25 |
| electrical | and hopefully build a kibana one as well | 14:25 |
| *** claudiub_ has joined #openstack-infra | 14:25 | |
| *** claudiub|3 has quit IRC | 14:25 | |
| fungi | hrm, yes we have one of each of those too ;) | 14:25 |
| *** claudiub|4 has quit IRC | 14:26 | |
| electrical | hehe | 14:26 |
| *** claudiub has quit IRC | 14:27 | |
| *** otter768 has joined #openstack-infra | 14:27 | |
| electrical | clarkb: feel free to ping me on IRC regarding the puppet modules whenever. I'm usually around :-) | 14:27 |
| *** wenlock has joined #openstack-infra | 14:28 | |
| *** afazekas has joined #openstack-infra | 14:29 | |
| *** craigbr has joined #openstack-infra | 14:29 | |
| openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck: Add query for osc 1.0.4 grenade bug 1454467 https://review.openstack.org/182696 | 14:29 |
| openstack | bug 1454467 in python-openstackclient "Latest version 1.0.4 doesn't work with master" [Undecided,New] https://launchpad.net/bugs/1454467 | 14:29 |
| *** chlong has quit IRC | 14:30 | |
| *** sks has quit IRC | 14:31 | |
| openstackgerrit | Merged openstack-infra/shade: Stop leaking server objects https://review.openstack.org/181513 | 14:31 |
| fungi | sometimes i feel like this channel exists solely so that bots can hold conversations with each other | 14:31 |
| openstackgerrit | Merged openstack-infra/puppet-zuul: Move dependencies into lib/ folder https://review.openstack.org/172007 | 14:31 |
| claudiub_ | fungi: hello. Quick question. Pushing a pre-release through gerrit (e.g.: 1.0.0a1) will not publish to pypi, right? Only releases. | 14:31 |
| electrical | fungi: haha yeah | 14:32 |
| *** otter768 has quit IRC | 14:32 | |
| fungi | claudiub_: it depends entirely on what jobs that repo has configured to run when. we do have pre-release jobs which usually upload wheels but no sdist tarballs | 14:32 |
| *** stevemar has joined #openstack-infra | 14:32 | |
| claudiub_ | fungi: I see. thanks for the answer. :) | 14:33 |
| *** carl_baldwin has joined #openstack-infra | 14:34 | |
| fungi | claudiub_: also the latest pbr release has some interesting consequences in our build/upload tooling with respect to pre-release tags. you probably want to use 1.0.0.0a1 but try building an sdist or wheel with the tag locally before you push it to make sure the version that pbr calculates matches the tag you used | 14:34 |
| fungi | we're hashing out ideas to cope with that in https://review.openstack.org/180660 | 14:35 |
| clarkb | the only logstash/elasticsearch module I could find that was useable was GPL licensed | 14:35 |
| clarkb | it was actually well done but incompatible for license reasons | 14:36 |
| clarkb | I do not know if this is the module being referred to above | 14:36 |
| clarkb | electrical ^ | 14:37 |
| electrical | clarkb: ours is apache licensed | 14:37 |
| *** deepakcs has joined #openstack-infra | 14:37 | |
| fungi | ahh, right, the whole "instantiating a class from a module in another module is tantamount to linking a library in gpl terms" | 14:37 |
| electrical | https://forge.puppetlabs.com/elasticsearch/elasticsearch | 14:37 |
| anteaya | clarkb: can you confirm the only gerrit db surgery you did yesterday was to stevebaker's row? | 14:37 |
| *** ivar-lazzaro has joined #openstack-infra | 14:38 | |
| clarkb | fungi at least thats tge theory | 14:38 |
| claudiub_ | fungi: I see. I'll take a look. doing a sdist locally looks fine. | 14:38 |
| clarkb | electrical then I dont think it existed at the time | 14:39 |
| electrical | clarkb: very possible yeah :-) | 14:39 |
| electrical | it used to exist under my own name but when i started working for Elastic i moved it | 14:39 |
| clarkb | also is the jenkins plugin the one jesusaurus wrote? thats neat | 14:42 |
| nibalizer | electrical: hiya | 14:42 |
| fungi | Shrews: so you have an e-mail showing a code review +1 from SpamapS on 181513 which was added at 05-12 19:15 utc, a minute after your ps5 upload, and it's not visible in the webui (that much i can confirm) | 14:42 |
| * nibalizer reads scrollback | 14:42 | |
| electrical | nibalizer: hiya mate :-) how are you doing? | 14:42 |
| clarkb | we dont use it because we don't redis and have a bunch of logs that need to be indexed that jenkins may not even know about | 14:42 |
| *** ddieterly has quit IRC | 14:42 | |
| clarkb | but it seemed to work well for jesusaurus | 14:42 |
| *** ivar-lazzaro has quit IRC | 14:43 | |
| electrical | clarkb: that's the one yeah | 14:43 |
| electrical | we just want to log the jenkins output. works fine for us :-) | 14:43 |
| electrical | and gives some of the nice junit numbers back as well | 14:43 |
| Shrews | fungi: yup! but, as i said, if the reason is not immediately known, don't worry about it unless you fear deep failure somewhere | 14:43 |
| mordred | fungi, clarkb: I'm not sure I agree with that theory, btw | 14:43 |
| clarkb | mordred ya its certainly debateable | 14:43 |
| mordred | fungi, clarkb: the gpl triggers on distribution, not use | 14:44 |
| *** mestery has quit IRC | 14:44 | |
| clarkb | mordred but iirc thats the Puppet theory | 14:44 |
| fungi | electrical: we basically stand up virtual cloud deployments and need to collect hundreds upon hundreds of discrete service logs, system logs and other data/configurations from each of them before we destroy them again, so yeah our jenkins is completely unaware of most of that | 14:44 |
| mordred | clarkb: the puppet theory is taht the GPL is a EULA? | 14:44 |
| electrical | fungi: ah yeah, makes sense yeah :-) | 14:44 |
| *** btully has joined #openstack-infra | 14:44 | |
| clarkb | mordred no that class inclusion is like linking | 14:45 |
| *** claudiub_ has quit IRC | 14:45 | |
| clarkb | mordred iirc puppet did a giant relicensing at one point as a result | 14:45 |
| mordred | clarkb: well, it's an interesting theory and I think it's based on wishful thinking on someone's part | 14:45 |
| fungi | mordred: basically that one library designing to the interface of another means that it becomes a derivative work | 14:45 |
| jeblair | mordred: sure, though i would like our software to be redistributable. | 14:45 |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Add query for osc 1.0.4 grenade bug 1454467 https://review.openstack.org/182696 | 14:45 |
| openstack | bug 1454467 in python-openstackclient "Latest version 1.0.4 doesn't work with master" [Undecided,New] https://launchpad.net/bugs/1454467 | 14:45 |
| mordred | jeblair: _that_ is a different thing altogether | 14:45 |
| nibalizer | electrical: im quite good, this is my 'home' project as it were | 14:46 |
| mordred | fungi: I believe that we do not want a world where APIs transfer copyright | 14:46 |
| jeblair | this is probably a better beer conversation | 14:46 |
| mordred | probably so | 14:46 |
| nibalizer | i'd love to move us to the official elastic modules (and rtyler/jenkins for that matter) but its hard and low on my list of priorities | 14:46 |
| fungi | agreed. we have lots to get done before this weekend | 14:46 |
| nibalizer | mordred: clarkb you should talk to electrical about beaker+docker since he led that effort | 14:46 |
| nibalizer | electrical: got a link to your jenkins handy? | 14:46 |
| mtreinish | fungi: it still doesn't feel like it's that soon | 14:47 |
| electrical | nibalizer: always :p http://build-eu-00.elastic.co/view/Puppet/ | 14:47 |
| fungi | mtreinish: i think my liver is already bracing for impact | 14:47 |
| *** bhunter71 has joined #openstack-infra | 14:48 | |
| clarkb | still no word from Gerrit on our upgrade fun that I can see :/ | 14:48 |
| *** fhubik has quit IRC | 14:48 | |
| *** craigbr has quit IRC | 14:48 | |
| *** mpaolino has joined #openstack-infra | 14:49 | |
| fungi | neither in their irc channel or ml | 14:50 |
| nibalizer | https://github.com/elastic/puppet-elasticsearch/blob/master/spec/spec_helper_acceptance.rb is where some of the 'magic' is, i'd like to see the pattern of s_h_a setting a proxy in the host become standard | 14:50 |
| *** mnaser has joined #openstack-infra | 14:50 | |
| *** e0ne is now known as e0ne_ | 14:50 | |
| mnaser | welp, fun mess today | 14:50 |
| fungi | mnaser: which one? that's not narrowing it down unfortunately | 14:51 |
| nibalizer | electrical: this is our current spec discussion https://review.openstack.org/#/c/178887/ | 14:51 |
| mnaser | fungi: http://venom.crowdstrike.com/ | 14:51 |
| *** samuelBartel has quit IRC | 14:51 | |
| fungi | mnaser: oh, that. i detest marketing names for security vulnerabilities | 14:51 |
| *** e0ne_ is now known as e0ne | 14:52 | |
| electrical | nibalizer: taking a look | 14:52 |
| mnaser | fungi: same. it looks like they're trying to use this as marketing material to promote their business | 14:52 |
| fungi | but i think it's cute that we still emulate floppies in virtual machines after all this time | 14:52 |
| mordred | nibalizer: isn't that something that could be put into a gem and required? | 14:52 |
| mnaser | fungi: the scary part is.. " It needs to be noted that even if a guest does not explicitly have a virtual floppy disk configured and attached, this issue is exploitable." | 14:52 |
| *** patrickeast has joined #openstack-infra | 14:52 | |
| mordred | nibalizer: like, I'd love it if as much boilerplate as is possible could be put into a gem or something so that we can just require 'openstack-infra-rspec' or something | 14:53 |
| *** hemna_ is now known as hemna | 14:53 | |
| fungi | mnaser: business promotion has basically been the ploy behind every one of the carefully-marketed vulnerabilities ever since heartbleed showed that it was a viable strategy and made it a trend | 14:53 |
| mnaser | fungi: yep, heartbleed really started a trend, heh | 14:54 |
| fungi | now they all need cute names or they didn't happen | 14:54 |
| mnaser | i think it's to get more online exposure by media, etc.. they love these cute names | 14:55 |
| mnaser | in the meantime, i have to figure out if we're making our own packages or hearing something from ubuntu | 14:55 |
| electrical | nibalizer: mordred clarkb If you guys need any info regarding beaker / docker / etc feel free to ping me. can always make time to have a chat about puppet module stuff | 14:55 |
| electrical | as in testing them :-) | 14:55 |
| nibalizer | mordred: yea that seems reasonable | 14:55 |
| mordred | electrical: awesome. thanks! our main interest around here is actually in running beaker without having beaker manager any vms or containers for us | 14:56 |
| mordred | electrical: but - conversations about testing puppet modules is definitely a thing we're excited to talk about | 14:56 |
| *** mase_x200 has quit IRC | 14:56 | |
| electrical | mordred: makes sense. there is a vmpooler thing so that beaker doesn't manage them. perhaps that's useful. | 14:57 |
| mordred | (although I could also see that it would be potentially nice for devs to be able to run the same tests locally with some docker) | 14:57 |
| clarkb | also how does anyone actually update their deps properly (but most existing work seems to only consume releases) | 14:57 |
| mordred | electrical: well, not really - we really don't want beaker to know anything about vms | 14:57 |
| clarkb | mordred: vmpooler is basically nodepool | 14:57 |
| mordred | electrical: we have this giant system for creating single use vms already | 14:57 |
| clarkb | mordred: but with redis and an api for devs to use instead of jenkins | 14:58 |
| mordred | and running workloads in side of those vms | 14:58 |
| nibalizer | electrical: http://git.openstack.org/cgit/stackforge/puppet-keystone/tree/spec/ check out that nodeset | 14:58 |
| *** Hal1 has quit IRC | 14:58 | |
| *** Hal1 has joined #openstack-infra | 14:58 | |
| mordred | so I think one of my goals would be that we help in whatever we can to make the experience of "hey beaker I already have a node and you're already on it" better | 14:58 |
| electrical | nibalizer: ah yeah. that uses vmpooler / nodepool thing right? | 14:58 |
| mordred | while at the same time making sure we enable devs who are not sitting inside of our CI system to be able to do testing ina sane manner | 14:59 |
| nibalizer | right now a node spins up, we install beaker on it, then beaker ssh's into localhost and runs puppet/test | 14:59 |
| electrical | nibalizer: ooh so, i see | 14:59 |
| *** spredzy is now known as spredzy|afk | 14:59 | |
| *** kmartin has joined #openstack-infra | 14:59 | |
| mordred | it would be neat if we could skip the ssh into localhost part :) | 14:59 |
| mordred | but it does work | 15:00 |
| electrical | mordred: https://github.com/puppetlabs/vmpooler something like that? | 15:00 |
| clarkb | mordred: we call that fork/exec | 15:00 |
| electrical | or you could write your custom hypervisor code :-) | 15:00 |
| mordred | electrical: hah. so - we have clouds | 15:00 |
| nibalizer | electrical: idunno if thats true.... | 15:00 |
| *** tiswanso has quit IRC | 15:00 | |
| nibalizer | electrical: its pretty tied to the ssh model | 15:00 |
| mordred | electrical: we don't want beaker to know about vmpooler/nodepool - that's the wrong layer of abstraction for us | 15:00 |
| *** reed_ has joined #openstack-infra | 15:00 | |
| nibalizer | ofc you know that codebase better than I do | 15:00 |
| electrical | hehe. whole beaker is fairly ssh focused yeah. | 15:01 |
| nibalizer | but they did allow for remotewindows support without ssh so its probably abstracted /somewhere/ | 15:01 |
| *** tiswanso_ has joined #openstack-infra | 15:01 | |
| nibalizer | but yea writing a hypervisor so that our beaker is a little less terrible is on my list | 15:01 |
| electrical | also serverspec / specinfra is ssh centric | 15:01 |
| nibalizer | as you can see, i have quite the list | 15:01 |
| mordred | it's possible that we may just have to live with ssh localhost for a while | 15:01 |
| *** asselin has joined #openstack-infra | 15:02 | |
| nibalizer | mordred: did you see my proposal to create a zuul job-doer that runs beaker and gets its nodes from nodepool? | 15:02 |
| jeblair | nibalizer: a zuulv3 one? | 15:02 |
| *** marcusvrn_ has joined #openstack-infra | 15:02 | |
| mordred | yah. that sounds like a zuul v3 thing | 15:02 |
| *** Sukhdev has joined #openstack-infra | 15:02 | |
| *** e0ne is now known as e0ne_ | 15:02 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add port resource methods https://review.openstack.org/179968 | 15:03 |
| *** BharatK has quit IRC | 15:03 | |
| openstackgerrit | Marton Kiss proposed openstack-infra/system-config: Disable puppet-askbot resources temporarily https://review.openstack.org/181894 | 15:04 |
| nibalizer | jeblair: yea it heads in the other direction from zuul v3 | 15:05 |
| mordred | nibalizer, electrical: the fundamental problem with having beaker manage vms is that we have a ton of jobs that are not puppet related. so we have to do all of the things that beaker wants to do with not-beaker as well. so beaker doing those things is anti-helpful. | 15:05 |
| *** ddieterly has joined #openstack-infra | 15:05 | |
| mordred | nibalizer, electrical: i do not want 20 different things that want to manage vms | 15:05 |
| mordred | I want one thing | 15:05 |
| nibalizer | mordred: well note that we could teach beaker (easily) to check a vm out from nodepool | 15:06 |
| mordred | nibalizer: we could - but why would we? | 15:06 |
| clarkb | nibalizer: thats still the wrong layer | 15:06 |
| *** scheuran has quit IRC | 15:06 | |
| nibalizer | im saying that one of the roles that ansible plays in zuulv3 could be done by beaker | 15:06 |
| mordred | nibalizer: it's the wrong layer of abstraction - we already know about workload to node relationships | 15:06 |
| dims | fungi: or any infra core: can you please add oslo-core to the oslo-cache-core in gerrit please (https://review.openstack.org/#/admin/groups/686,members) | 15:06 |
| mordred | nibalizer: yes. we could chose to use beaker instead of ansible | 15:06 |
| mordred | but we will not chose to do that | 15:06 |
| openstackgerrit | Marton Kiss proposed openstack-infra/puppet-askbot: Askbot module refactor https://review.openstack.org/181873 | 15:07 |
| mordred | because ansible is more generally applicable to the task of "run this arbitrary workload on this node" and beaker is a tool for testing puppet modules | 15:07 |
| mordred | testing puppet modules is only one of the things we do | 15:07 |
| mordred | whereas ansible can easily express _All_ of the thigns we do, including testing puppet modules | 15:07 |
| nibalizer | mordred: the idea though that there is code running outside the system-under-test that does things to the system-under-test is a good idea | 15:07 |
| mordred | nibalizer: isolation is a good thing - but we need to solve it once generally and not once per workload type | 15:08 |
| *** erikmwilson has quit IRC | 15:09 | |
| *** aduarte has quit IRC | 15:09 | |
| *** unicell1 has joined #openstack-infra | 15:10 | |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Add glance_store to projects.txt https://review.openstack.org/182681 | 15:11 |
| nibalizer | mordred: sure but that puts us back in the spot where we are now | 15:11 |
| *** erikmwilson has joined #openstack-infra | 15:11 | |
| mordred | nibalizer: right. where I want the thing that runs tests to be able to run on the host I tell it to run on | 15:11 |
| mordred | nibalizer: and I'd liek to request that it not try to manage the lifecycle of such things | 15:12 |
| *** unicell has quit IRC | 15:12 | |
| mordred | nibalizer: if I want to do a 2-node test - workload on one box connecting to second box that is being tested | 15:12 |
| electrical | mordred: you can supply a host and login to run it on in the nodeset config. not sure if you could do it via env vars or something else | 15:12 |
| mordred | nibalizer: I can already express the 2-node setup | 15:12 |
| *** juggler has quit IRC | 15:13 | |
| *** panda has quit IRC | 15:13 | |
| mordred | electrical: yah. I _think_ that's what we're doing now, right? | 15:13 |
| electrical | mordred: yeah. without setting a hypervisor and setting an IP you tell it where to connect to. | 15:14 |
| *** nelsnelson has joined #openstack-infra | 15:14 | |
| electrical | so you will have an other process getting the VM you want to test on and passing it to beaker | 15:14 |
| *** bswartz has quit IRC | 15:14 | |
| mordred | right. although it's, for us, always going to be "localhost" | 15:14 |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Update ci for F21 jenkins nodes https://review.openstack.org/182219 | 15:14 |
| mordred | at least until we get to the point where having more than one host involved makes sense | 15:15 |
| *** panda has joined #openstack-infra | 15:15 | |
| *** rlandy has quit IRC | 15:15 | |
| *** arielb has joined #openstack-infra | 15:15 | |
| *** arielb has joined #openstack-infra | 15:15 | |
| *** dims has quit IRC | 15:15 | |
| clarkb | well you also have to specify the distro even if no hypervisor is involved | 15:15 |
| clarkb | and Linux isn't specific enough apparently | 15:16 |
| electrical | ah yeah very true | 15:16 |
| nibalizer | idunno how jira works really: https://tickets.puppetlabs.com/browse/BKR/?selectedTab=com.atlassian.jira.jira-projects-plugin:issues-panel | 15:16 |
| clarkb | given it uses vagrant I was hoping that would be useable | 15:16 |
| *** dims has joined #openstack-infra | 15:16 | |
| nibalizer | clarkb: that should be better once we write our own hypervisor | 15:16 |
| *** matrohon has quit IRC | 15:16 | |
| pelix | waynr: mainly yolanda 's work, I just helped describe the size of the impact ;) | 15:16 |
| mordred | nibalizer: so - when you say "write our own hypervisor" | 15:17 |
| *** spredzy|afk is now known as spredzy | 15:18 | |
| yolanda | pelix, you helped a lot also in making that code more elegant and robust | 15:18 |
| mordred | nibalizer: you mean "write a beaker driver that works like we want" right | 15:18 |
| nibalizer | mordred: yep | 15:18 |
| mordred | nibalizer: you do not mean write our own hypervisor like write our own hypervisor | 15:18 |
| nibalizer | https://github.com/nibalizer/beaker-libvirt/blob/master/lib/beaker/hypervisor/libvirt.rb libvirt example | 15:18 |
| mordred | ok. | 15:18 |
| mordred | good | 15:18 |
| *** mrmartin has joined #openstack-infra | 15:18 | |
| mordred | so we'd write a no-op hypervisor that does not provision things and returns localhost as the IP? | 15:18 |
| nibalizer | mordred: ofc I'm not tracking what greghaynes is doing, so that could be in the works, idunno | 15:19 |
| clarkb | mordred: no he means make beaker understand a thing they call hypervisor which may be cloud or just fork | 15:19 |
| nibalizer | mordred: yup | 15:19 |
| mordred | +100 | 15:19 |
| nibalizer | clarkb: i dont think we can just fork | 15:19 |
| *** jistr is now known as jistr|mtg | 15:19 | |
| clarkb | nibalizer: thats unfortunate | 15:19 |
| nibalizer | thats not trivial anyways | 15:19 |
| *** bswartz has joined #openstack-infra | 15:19 | |
| nibalizer | its really tied to returning connection information | 15:19 |
| clarkb | should be possible to fork with a pipe/socket between parent child then have child run puppet things | 15:20 |
| waynr | d'oh! sorry yolanda, nice work both of you :) | 15:20 |
| clarkb | ya so you can still do that with a fork | 15:20 |
| clarkb | but instead of tcp its unix | 15:20 |
| *** sabeen has joined #openstack-infra | 15:21 | |
| *** dkranz has joined #openstack-infra | 15:21 | |
| *** scheuran has joined #openstack-infra | 15:21 | |
| yolanda | waynr, no problem for the credits, if jjb is faster i'm happy, it could be pelix, zaro, you , whoever | 15:22 |
| fungi | dims: done | 15:22 |
| dims | fungi: awesome thanks! | 15:22 |
| openstackgerrit | Merged openstack-infra/pypi-mirror: If we specify a branch, always use it https://review.openstack.org/182608 | 15:22 |
| *** salv-orlando has joined #openstack-infra | 15:22 | |
| electrical | clarkb: im staying out further of the discussion how to connect to the VM's. i can help though with any other puppet module testing questions ;-) | 15:23 |
| electrical | or docker related stuff with it | 15:23 |
| *** afazekas_ has quit IRC | 15:24 | |
| *** afazekas has quit IRC | 15:24 | |
| openstackgerrit | Julien Danjou proposed openstack-infra/git-review: Allow to specify which commit to send https://review.openstack.org/139701 | 15:26 |
| *** jistr|mtg is now known as jistr | 15:29 | |
| anteaya | I've lost track, changing the commit message is considered a new patch and purposely zeros out prior votes, yes? https://review.openstack.org/#/c/169789/ | 15:29 |
| *** sdake has quit IRC | 15:30 | |
| jeblair | anteaya: yes | 15:30 |
| anteaya | great thanks for participating in my reality check | 15:30 |
| *** sdake has joined #openstack-infra | 15:30 | |
| clarkb | ya we can have it carry votes forward but we put depends-on in commit messages and some tests check commit message content | 15:31 |
| jeblair | (and some people -1 on commit msg content) | 15:31 |
| anteaya | true | 15:32 |
| anteaya | just ensuring my rememberances of reality sync up with current reality | 15:32 |
| anteaya | I sometimes get folds in the space-time continum | 15:33 |
| *** yamahata has joined #openstack-infra | 15:33 | |
| *** ildikov has quit IRC | 15:33 | |
| clarkb | dhellmann: sdague IIRC you were looking for a python3.4 compat openstack job for running under devstack with python3.4. Looks like Ironic's unittests pass python3.4 now at least, possible that it is a good option to be the devstack python3.4 guinea pig? | 15:35 |
| *** yamahata has quit IRC | 15:36 | |
| *** yamahata has joined #openstack-infra | 15:36 | |
| *** dangers_away is now known as dangers | 15:37 | |
| *** mriedem has quit IRC | 15:37 | |
| openstackgerrit | Derek Higgins proposed openstack-infra/tripleo-ci: Allow puppet jobs to also run on F21 https://review.openstack.org/182722 | 15:37 |
| *** mriedem has joined #openstack-infra | 15:37 | |
| mtreinish | fungi, clarkb, mordred: so it looks like some of the stable issues are being caused by pycadf 0.6.1 from last week not actually making it to pypi | 15:38 |
| mordred | hrm | 15:38 |
| mtreinish | any chance you guys could kick off the pypi job on that 0.6.1 tag again? | 15:38 |
| mriedem | true story https://pypi.python.org/pypi/pycadf/0.6.1 | 15:39 |
| *** ivar-lazzaro has joined #openstack-infra | 15:39 | |
| gordc | mtreinish: just came to mention that | 15:39 |
| *** spredzy is now known as spredzy|afk | 15:39 | |
| mriedem | gordc: we're nerding it up in #openstack-stable btw | 15:39 |
| gordc | mriedem: thanks. i'll join | 15:39 |
| clarkb | http://logs.openstack.org/a7/a704daf4d64f5d13c1419cdcce36b2c04916c613/release/ | 15:40 |
| clarkb | does pycadf have a pypi release job? /me goes to look | 15:40 |
| clarkb | appears to | 15:40 |
| gordc | clarkb: it does... i'm guessing it failed at the publish task when i released last week | 15:41 |
| mtreinish | clarkb: yeah it does: http://git.openstack.org/cgit/openstack-infra/project-config/tree/zuul/layout.yaml#n2054 | 15:41 |
| gordc | http://tarballs.openstack.org/pycadf/ | 15:41 |
| *** salv-orlando has quit IRC | 15:41 | |
| clarkb | hrm it made sdist and whl but no job log for it | 15:41 |
| mordred | maybe it happened to happen at just the right time to catch the wrong part of a system restart | 15:42 |
| fungi | i wonder if the log upload failed, and so the job didn't succeed thus no pypi upload job ran | 15:42 |
| mordred | or that | 15:42 |
| clarkb | fungi: thats possible | 15:42 |
| *** sabeen2 has joined #openstack-infra | 15:42 | |
| clarkb | so two ways we can fix this, we can rerun the pypi upload job or gordc can upload to pypi | 15:42 |
| fungi | which would explain why 1. no log and 2. no pypi upload | 15:43 |
| *** e0ne_ is now known as e0ne | 15:43 | |
| *** Krinkle is now known as Krinkle|detached | 15:43 | |
| mordred | re-running the pypi upload job will ensure that the tarball in tarballs.o.o is the same as the one in pypi | 15:43 |
| clarkb | mordred: well gordc should use the tarball on tarballs.o.o but ya | 15:43 |
| *** sabeen has quit IRC | 15:43 | |
| clarkb | I can trigger the pypi upload job | 15:43 |
| *** ivar-lazzaro has quit IRC | 15:44 | |
| gordc | clarkb: cool cool. works for me. | 15:44 |
| fungi | i'm checking zuul logs from the 5th | 15:44 |
| *** gyee has joined #openstack-infra | 15:44 | |
| *** eharney has joined #openstack-infra | 15:45 | |
| fungi | theory disproven... 2015-05-05 14:26:41,516 INFO zuul.Gearman: Build <gear.Job 0x7fb76558ba50 handle: H:127.0.0.1:1876823 name: build:pycadf-tarball:bare-trusty unique: be7d9522939b43718b2a01600a5ab45d> complete, result SUCCESS | 15:45 |
| mordred | oh. well gross | 15:45 |
| fungi | 2015-05-05 16:34:52,048 INFO zuul.Gearman: Gearman job <gear.Job 0x7fb7a1bdfe50 handle: H:127.0.0.1:1877647 name: build:pycadf-pypi-both-upload unique: c551987d85e44f5cac8da0a1b6124efc> lost due to disconnect | 15:46 |
| *** zz_ja is now known as ja | 15:46 | |
| mtreinish | oh, that's something | 15:46 |
| clarkb | fungi: so its possible the console log upload thread went away | 15:47 |
| fungi | no job history at all for https://jenkins.openstack.org/job/pycadf-pypi-both-upload/ | 15:47 |
| *** kgiusti has quit IRC | 15:47 | |
| clarkb | fungi I am ready to retrigger the job if you are | 15:47 |
| fungi | go for it | 15:47 |
| fungi | looks like it's running now | 15:48 |
| fungi | failed | 15:48 |
| clarkb | https://jenkins.openstack.org/job/pycadf-pypi-both-upload/1/console | 15:48 |
| fungi | HTTPError: 400 Client Error: This filename has previously been used, you should use a different version. | 15:48 |
| fungi | saywhat? | 15:48 |
| clarkb | I think thats pypi telling us it alrady has the file | 15:48 |
| *** banix has joined #openstack-infra | 15:49 | |
| gordc | fungi: argh... that might have been me | 15:49 |
| clarkb | https://pypi.python.org/simple/pycadf/ except not | 15:49 |
| fungi | gordc: yeah, pypi will let you delete a file, but won't let you reupload a file with the same name as one uploaded in the past | 15:49 |
| fungi | gordc: so if you uploaded and then deleted it, that would be an explanation | 15:50 |
| *** tim_o has joined #openstack-infra | 15:50 | |
| gordc | fungi: that's some good rules... :/ | 15:50 |
| gordc | fungi: i guess we need to move to 0.6.2? | 15:50 |
| fungi | gordc: that would be simplest | 15:50 |
| fungi | re-tag 0.6.1 as 0.6.2 should work fine | 15:51 |
| *** Krinkle|detached is now known as Krinkle | 15:51 | |
| gordc | fungi: ok. i'll grab the right sha | 15:51 |
| gordc | apologies on the madness | 15:51 |
| fungi | it happens | 15:52 |
| gordc | i should probably delete the empty 0.6.1 listing on pypi since it's effectively useless | 15:52 |
| fungi | yeah, would be safest | 15:52 |
| clarkb | did the delete happen such that the original job may have gone sideways due to that | 15:52 |
| *** esker has joined #openstack-infra | 15:53 | |
| fungi | that's what it sounded like anyway | 15:53 |
| *** Krinkle is now known as Krinkle|detached | 15:55 | |
| gordc | ./release_library.sh juno 0.6.2 b9c1d41b83e2f5937a69070df8dd7c12439e333a pycadf | 15:57 |
| gordc | ^ seem sane? | 15:57 |
| clarkb | gordc: we don't use release_library around here so I dunno | 15:57 |
| clarkb | but I can check the sha1 in just a moment | 15:57 |
| *** dprince has quit IRC | 15:58 | |
| *** ashleighfarnham has joined #openstack-infra | 15:58 | |
| clarkb | gordc: b9c1d41b83e2f5937a69070df8dd7c12439e333a is the commit that 0.6.1 points to so that lgtm | 15:58 |
| gordc | clarkb: cool cool. i'll confirm the release_library stuff... will push once i've done that. | 15:59 |
| *** devvesa has quit IRC | 15:59 | |
| *** markus_z has quit IRC | 16:00 | |
| fungi | yeah, i think that's dhellmann's tool which generates the release notifications and stuff too. haven't really looked at it | 16:00 |
| *** nfedotov has quit IRC | 16:02 | |
| *** arielb has quit IRC | 16:04 | |
| gordc | clarkb: fungi: cool cool i've kicked off another job. for 0.6.2 | 16:04 |
| electrical | uhg. co-worker is not seeing the advantages of using jenkins job builder thing :-( | 16:05 |
| *** ashleighfarnham has quit IRC | 16:06 | |
| *** unicell1 has quit IRC | 16:06 | |
| fungi | electrical: templating engine, also easier to write/edit/review/maintain yaml in a revision control system than xml | 16:06 |
| *** markvan has quit IRC | 16:06 | |
| *** alexsyip has joined #openstack-infra | 16:07 | |
| *** jlanoux has quit IRC | 16:07 | |
| mriedem | electrical: tell them to own all of the jenkins job configs in your CI then and have fun doing copy/paste everywhere | 16:07 |
| mriedem | electrical: maybe that's the point - job security :) | 16:07 |
| electrical | mriedem: haha :p | 16:07 |
| electrical | he doesn't even maintains the jobs :p | 16:07 |
| *** ja is now known as zz_ja | 16:07 | |
| mriedem | everyone has to go through bob to get jenkins changes rather than help themselves to add new configs or update jobs | 16:07 |
| mriedem | then ignore bob | 16:07 |
| electrical | hehe yeah | 16:08 |
| *** Swami has joined #openstack-infra | 16:08 | |
| *** arielb has joined #openstack-infra | 16:08 | |
| *** jcoufal has quit IRC | 16:08 | |
| *** markvan has joined #openstack-infra | 16:08 | |
| electrical | To be fair, i don't really care if he wants to use it or not, i'll defo use it | 16:09 |
| electrical | got my account and api key so i can manage them anyway | 16:09 |
| mordred | \o/ | 16:09 |
| electrical | :p | 16:10 |
| mriedem | mrodden: is much happier that we use jjb and i don't have to bug him for job changes internally | 16:10 |
| *** tiswanso_ has quit IRC | 16:10 | |
| *** abregman has joined #openstack-infra | 16:10 | |
| mriedem | i just have to bug him to review the changes | 16:10 |
| electrical | hehe i can imagine :-) | 16:10 |
| electrical | i fully maintain my own jobs anyway :-) | 16:10 |
| *** tiswanso has joined #openstack-infra | 16:11 | |
| gordc | mriedem: just an fyi, pycadf 0.6.2 is released and on pypi | 16:12 |
| *** erlon has joined #openstack-infra | 16:12 | |
| *** soren has quit IRC | 16:12 | |
| *** rwsu has joined #openstack-infra | 16:13 | |
| mriedem | gordc: sweet | 16:13 |
| greghaynes | oh hey, https://review.openstack.org/#/c/168603/ is 2x +2 | 16:13 |
| greghaynes | clarkb: fungi ^ when youall want to try out the first nodepool-shade bits? | 16:14 |
| clarkb | greghaynes: mordred wanted that other change in first which I pulled up to review then got distracted. Let me do that now | 16:14 |
| clarkb | then I am assuming mordred wants to release? then we shade nodepool? | 16:14 |
| greghaynes | clarkb: oh? the cacert one? | 16:14 |
| clarkb | greghaynes: a | 16:14 |
| greghaynes | clarkb: b | 16:14 |
| clarkb | *ya :P | 16:14 |
| greghaynes | clarkb: https://review.openstack.org/#/c/182488/ | 16:14 |
| fungi | clarkb: mordred: greghaynes: oh, right, we needed to tag a nodepool release which includes dib support. not sure if that should be pre-shade or post-shade | 16:15 |
| clarkb | fungi: we should robably make a tag pre shade but that can happen after shade code lands I think | 16:16 |
| greghaynes | ah, does the puppet code only deploy releases? | 16:16 |
| *** amitgandhinz has quit IRC | 16:16 | |
| clarkb | greghaynes: not ours | 16:16 |
| clarkb | greghaynes: but for other consumers | 16:16 |
| greghaynes | ok | 16:16 |
| clarkb | noted a style thing inline but +2, can approve if no one else thinks that change should be made | 16:18 |
| clarkb | mordred: greghaynes ^ | 16:18 |
| dougwig | good morning, infra. what is the procedure for a repo rename? https://review.openstack.org/#/c/179417/ | 16:19 |
| greghaynes | meh, id say just merge it | 16:19 |
| mordred | yah - we can get that later | 16:19 |
| mordred | clarkb: also - yeah - we should tag before we add the shade dependency I think | 16:19 |
| * clarkb approves | 16:19 | |
| mordred | clarkb: so that folks using via packages have a good pausing point before they have to package shade too | 16:20 |
| clarkb | dougwig: this is a move from stackforge/octavia to openstack/octavia? | 16:20 |
| dougwig | clarkb: yes. | 16:20 |
| *** ashleighfarnham has joined #openstack-infra | 16:20 | |
| *** scheuran has quit IRC | 16:21 | |
| *** ashleighfarnham has quit IRC | 16:21 | |
| clarkb | dougwig: it basically looks like a change that s/stackforge/openstack/ for octavia in openstack-infra/project-config but it requires a gerrit downtime so you want to also add yourself to the list at https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting so we know which ones are in the pipeline | 16:21 |
| clarkb | dougwig: its looking like next batch of renameswill happen post summit | 16:21 |
| *** ashleighfarnham has joined #openstack-infra | 16:22 | |
| dougwig | clarkb: ok, thanks. is there a redirect in place after, by any chance? | 16:22 |
| clarkb | dougwig: there is not | 16:22 |
| dougwig | clarkb: ok, we'll mention both urls and the impending rename at the summit then, no worries. thanks, i'll get the review up today. | 16:23 |
| *** aduarte has joined #openstack-infra | 16:25 | |
| *** jistr has quit IRC | 16:25 | |
| *** Krinkle|detached is now known as Krinkle | 16:26 | |
| *** otter768 has joined #openstack-infra | 16:28 | |
| *** dprince has joined #openstack-infra | 16:29 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add port resource methods https://review.openstack.org/179968 | 16:29 |
| openstackgerrit | Merged openstack/requirements: Add appdirs requirement https://review.openstack.org/182061 | 16:30 |
| openstackgerrit | Merged openstack/requirements: Add appdirs requirement https://review.openstack.org/182061 | 16:30 |
| mordred | wow. it merged twice did it? | 16:30 |
| clarkb | I think sometimes gerrit gets confused and emits two events | 16:30 |
| jeblair | mordred: we do that for the really important changes | 16:30 |
| mordred | jeblair: cool. we _SHOULD_ do that | 16:31 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack-dev/hacking: Updated from global requirements https://review.openstack.org/181908 | 16:31 |
| *** dannywilson has joined #openstack-infra | 16:31 | |
| *** otter768 has quit IRC | 16:32 | |
| *** dizquierdo has quit IRC | 16:34 | |
| *** kgiusti has joined #openstack-infra | 16:34 | |
| clarkb | mordred: jeblair fungi pleia2 jhesketh SergeyLukjanov should we go ahead and tag nodepool on the current commit then? | 16:34 |
| clarkb | then merge the use shade change and restart nodepool some time today? | 16:34 |
| clarkb | I should check if lxml can build on nodepool.o.o | 16:34 |
| *** ihrachyshka has quit IRC | 16:34 | |
| *** yfried__ has quit IRC | 16:35 | |
| *** tonytan4ever has quit IRC | 16:35 | |
| mordred | clarkb: ++ | 16:35 |
| mordred | clarkb: good call | 16:35 |
| greghaynes | SGTM | 16:35 |
| clarkb | python-lxml is installed via apt | 16:36 |
| clarkb | so we should be fine | 16:36 |
| mordred | btw. I just cut 1.0.0 of os-client-config | 16:36 |
| greghaynes | wooo! | 16:37 |
| *** derekh_ has quit IRC | 16:37 | |
| greghaynes | that means its done, right/ | 16:37 |
| clarkb | yes | 16:37 |
| pleia2 | clarkb: thumbs up | 16:38 |
| openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: Rename stackforge/octavia -> openstack/octavia https://review.openstack.org/182748 | 16:38 |
| *** unicell has joined #openstack-infra | 16:39 | |
| *** ivar-lazzaro has joined #openstack-infra | 16:40 | |
| openstackgerrit | Mikhail S Medvedev proposed openstack-infra/puppet-zuul: Create lib/ folder and change respective requires https://review.openstack.org/182750 | 16:40 |
| *** erikmwilson has quit IRC | 16:40 | |
| *** erikmwilson has joined #openstack-infra | 16:40 | |
| zaro | morning | 16:40 |
| clarkb | make an 0.2.0 nodepool tag | 16:40 |
| clarkb | er make the tag on nodepool 0.2.0? | 16:40 |
| anteaya | morning zaro | 16:41 |
| zaro | linuxmaniac: been busy with gerrit lately, will take a look at that change again soon. | 16:41 |
| openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: Rename stackforge/octavia -> openstack/octavia https://review.openstack.org/182748 | 16:41 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Consume os_client_config defaults as base defaults https://review.openstack.org/181652 | 16:41 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Rely on defaults being present https://review.openstack.org/181653 | 16:41 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Make ironic use the API version system https://review.openstack.org/181651 | 16:41 |
| *** mpaolino has quit IRC | 16:42 | |
| *** soren_ has joined #openstack-infra | 16:42 | |
| *** jlanoux has joined #openstack-infra | 16:42 | |
| openstackgerrit | Mikhail S Medvedev proposed openstack-infra/puppet-zuul: Create lib/ folder and change respective requires https://review.openstack.org/182750 | 16:43 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: Experimental job for Nova API services under Apache2 https://review.openstack.org/182753 | 16:44 |
| *** prad has quit IRC | 16:44 | |
| *** ivar-lazzaro has quit IRC | 16:44 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: [WIP] Experimental job for Nova API services under Apache2 https://review.openstack.org/182753 | 16:44 |
| *** prad has joined #openstack-infra | 16:44 | |
| clarkb | jeblair: fungi thoughts on a 0.2.0 nodepool tag based on current HEAD? | 16:44 |
| openstackgerrit | Merged openstack-infra/shade: Don't error on missing certs https://review.openstack.org/182488 | 16:45 |
| *** dimtruck is now known as zz_dimtruck | 16:45 | |
| fungi | clarkb: that seems reasonable | 16:45 |
| *** marun has joined #openstack-infra | 16:46 | |
| fungi | hashar: did you have any packaging preferences on tag version and which commit for nodepool? | 16:46 |
| fungi | (if you happen to be around) | 16:46 |
| *** claudiub has joined #openstack-infra | 16:46 | |
| *** amotoki_ has quit IRC | 16:46 | |
| clarkb | fungi: oh good uestion since hashar likely is the biggest ocnsumer of those tags | 16:47 |
| jeblair | clarkb, mordred: what's the delta since the last one? | 16:47 |
| clarkb | jeblair: its actually pretty small, some docs updates, some bug fixes, some logging changes | 16:48 |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Add a note about the order of pycadf and Babel because of pytz https://review.openstack.org/182755 | 16:48 |
| zaro | sdague: do you have a suggestion with regards to the upstream gerrit query fix? https://gerrit-review.googlesource.com/#/c/62880 | 16:48 |
| jeblair | clarkb: should it be a 0.1.1? | 16:48 |
| clarkb | jeblair: let me read the diff more carefully, but that is very possible | 16:49 |
| fungi | clarkb: jeblair oh, i missed we grew a 0.1.0 for some reason. 7 non-merge commits | 16:49 |
| zaro | sdague: i gave up because I just wasn't getting any help from upstream cores. | 16:49 |
| jeblair | yeah, the 0.1.0 was the "pre-shade" tag that hashar asked for | 16:49 |
| hashar | fungi: clarkb: jeblair: yeah tags when you are breaking dependencies or adding new breaking features would be nice | 16:49 |
| fungi | rather, i missed that 0.1.0 was recent | 16:49 |
| hashar | makes it easier to catch up with your changes :] | 16:50 |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Fix path issue to controller-config-pacemaker https://review.openstack.org/182553 | 16:50 |
| fungi | so maybe we just tag the commit that drags in shadeing in that case? | 16:50 |
| jeblair | my read of the changelog looks more like a 0.1.1 which i think is a swell idea | 16:50 |
| fungi | that works too | 16:50 |
| clarkb | jeblair: yup mine too | 16:50 |
| clarkb | 0.1.1 is what it should be | 16:50 |
| jeblair | clarkb: all yours if you want it :) | 16:51 |
| clarkb | jeblair: doing | 16:51 |
| *** patrickeast_ has joined #openstack-infra | 16:51 | |
| *** patrickeast has quit IRC | 16:52 | |
| *** patrickeast_ is now known as patrickeast | 16:52 | |
| zaro | pelix, electrical : do we have to fix something in python-jenkins for python 3.x | 16:53 |
| *** zz_dimtruck is now known as dimtruck | 16:53 | |
| *** shashankhegde has joined #openstack-infra | 16:54 | |
| electrical | zaro: yeah. the http call needs to do a decode to utf8. i have to run but pelix or fungi have a snippet of the fix | 16:54 |
| *** signed8bit has joined #openstack-infra | 16:55 | |
| *** Krinkle is now known as Krinkle|detached | 16:55 | |
| anteaya | so zuul's git refs show the latest zuul tag at 2.0.0: http://git.openstack.org/cgit/openstack-infra/zuul/refs/ somehow patrickeast's zuul version is 2.0.1.dev392: http://paste.openstack.org/show/221977/ | 16:55 |
| anteaya | can anyone help me understand where zuul 2.0.1 is coming from? | 16:55 |
| clarkb | tag pushed | 16:55 |
| *** Krinkle|detached is now known as Krinkle | 16:56 | |
| clarkb | anteaya: pbr 0.11 increments the version for you based on semver | 16:56 |
| jlvillal | The openstack/python-ironicclient gate is blocked by this patch: https://review.openstack.org/#/c/182340/ Have to blacklist httpretty. | 16:56 |
| anteaya | clarkb: ah | 16:56 |
| jlvillal | Reviews appreciated. | 16:56 |
| greghaynes | clarkb: awesome, and so when we merge the nodepool change do we have to manually restart nodepool or is that an automagic rollout? | 16:56 |
| clarkb | anteaya: so the next version of zuul will be 2.0.1 and the intermediate commits are the appended .devXXX | 16:56 |
| anteaya | clarkb: that was the piece I was missing thanks | 16:56 |
| clarkb | greghaynes: we have to manually restart nodepool | 16:56 |
| anteaya | okay thanks | 16:56 |
| greghaynes | fun | 16:56 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/requirements: Pin os-brick to the current release https://review.openstack.org/182756 | 16:57 |
| clarkb | jlvillal: I am pretty sure I said next time we got one of those patches I would remove httpretty from requirements... | 16:57 |
| clarkb | jlvillal: is there a good reason to continue using a lib that we constantly run into problems with? | 16:58 |
| fungi | clarkb: anteaya: well, the next version of zuul will be _at_least_ 2.0.1 but might be higher | 16:58 |
| jlvillal | clarkb: If that happens, can we keep using it. I'm not sure the work involved in removing it. | 16:58 |
| *** tiswanso has quit IRC | 16:58 | |
| clarkb | jlvillal: I am not either, but I do know that ever few months we exclude the next httpretty release for some reason instead of A) fixing our tests or B) replacing the lib | 16:58 |
| jlvillal | clarkb: I guess the good reason is the work required to remove it from all the unittests. I have no idea the effort involved. | 16:58 |
| anteaya | fungi: yes, so 2.0.1 is pbr's best guess at a minimal increment based on the existing tag of 2.0.0 | 16:59 |
| *** tiswanso has joined #openstack-infra | 16:59 | |
| pelix | zaro: seems urlopen on python 3 returns a unicode byte string, while json.loads() expects a string. So should be calling '.decode("utf-8")' on the read() before passing it to json.loads(..) | 16:59 |
| pelix | zaro: I expect it would be a good idea to modify the tests to expose this on python 3 first before adding the fix, since I suspect the tests are working around the problem | 17:00 |
| *** sdake has quit IRC | 17:00 | |
| fungi | anteaya: exactly | 17:00 |
| *** jcoufal has joined #openstack-infra | 17:01 | |
| jlvillal | clarkb: I am mentioning the issue over in the #openstack-ironic channel. To see if we can add it to an agenda item to discuss. | 17:01 |
| *** asselin has quit IRC | 17:01 | |
| fungi | anteaya: and "dev" means "sort before" (roughly) | 17:01 |
| jlvillal | clarkb: Do you know how many projects use httpretty? Is there an easy way to find out? | 17:02 |
| clarkb | jlvillal: apparently there has been a patch to upstream since february 20th | 17:02 |
| jlvillal | clarkb: Yes :( | 17:02 |
| clarkb | jlvillal: there really isn't an easier way than cloning all of them and grepping for httpretty | 17:02 |
| greghaynes | clarkb: looks like the httpretty author suggests not using pip :p | 17:02 |
| *** Sukhdev has quit IRC | 17:02 | |
| *** ssam2 has quit IRC | 17:02 | |
| greghaynes | heh | 17:02 |
| greghaynes | (reading their README) | 17:02 |
| *** jcoufal has quit IRC | 17:02 | |
| *** deepakcs has quit IRC | 17:03 | |
| *** Sukhdev has joined #openstack-infra | 17:03 | |
| clarkb | jlvillal: commented on the requirements chnage | 17:03 |
| anteaya | fungi: thanks | 17:03 |
| *** patrickeast has quit IRC | 17:03 | |
| *** spredzy|afk is now known as spredzy | 17:03 | |
| *** mrmartin has quit IRC | 17:03 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: [WIP] Experimental job for Nova API services under Apache2 https://review.openstack.org/182753 | 17:03 |
| *** hashar has quit IRC | 17:03 | |
| jlvillal | clarkb: Thanks | 17:04 |
| *** davideag_ has joined #openstack-infra | 17:04 | |
| *** sks has joined #openstack-infra | 17:05 | |
| clarkb | mordred: do you need to do anything else shade side (tag maybe?) before we approve the nodepool change? | 17:06 |
| *** sks has quit IRC | 17:06 | |
| fungi | clarkb: jlvillal: cloudcafe, python-surveilclient, surveil, python-openstacksdk, cloud-init*, ospurge, python-group-based-policy-client, python-ironicclient, congress* (*requires but does not import) | 17:07 |
| *** davideagnello has quit IRC | 17:07 | |
| jlvillal | fungi: Thanks. And a fair amount of projects... | 17:07 |
| fungi | also quite a few projects include changelog entries about how they've removed it because <problem> | 17:08 |
| clarkb | anyways I thinkwe should just require <0.8.7 for now, then when upstream fixes we can swing around nd loosen the requirement | 17:08 |
| fungi | so looks like 7 projects in gerrit actively use this, most of which are not actually using openstack/requirements | 17:09 |
| fungi | would be useful to cross-reference that short-list against who's running requirements sync and enforcement | 17:09 |
| fungi | my guess is somewhere between 2-4 | 17:09 |
| *** bhuvan-afk has joined #openstack-infra | 17:10 | |
| *** shashankhegde has quit IRC | 17:11 | |
| mordred | clarkb: I don't think there is anything outstanding that would affect nodepool... lemme check | 17:11 |
| openstackgerrit | Samuel de Medeiros Queiroz proposed openstack-infra/project-config: Experimental Devstack job for Identity v3 API only https://review.openstack.org/179661 | 17:12 |
| *** davideagnello has joined #openstack-infra | 17:12 | |
| clarkb | but once we get the nodepool shade change in I want to flip over to figuring out this zuul-cloner + dib git cache interaction | 17:13 |
| mordred | clarkb: let me tag current state though - there was a small requirements change and I'd rather make puppet install it all right the first time because *lame* | 17:13 |
| *** claudiub|2 has joined #openstack-infra | 17:13 | |
| *** achanda has joined #openstack-infra | 17:14 | |
| anteaya | samueldmq: the comment in patchset 4 asked you to include a depends-on in the commit message: https://review.openstack.org/#/c/179661/ | 17:14 |
| *** claudiub has quit IRC | 17:14 | |
| mordred | clarkb: k. 0.6.1 pushed - we should be good to go when it hits the pypi | 17:14 |
| *** HeOS has quit IRC | 17:15 | |
| *** davideag_ has quit IRC | 17:15 | |
| fungi | okay, we're past successfully building/uploading centos-6 dib images in hpcloud now, but when they boot nodepool can't ssh to them... what's the easiest way to debug? nova console? | 17:16 |
| *** pelix has quit IRC | 17:16 | |
| clarkb | fungi: I would boot one off the same image | 17:16 |
| clarkb | fungi: if I had to venture a guess we actually do need cloud-init for something | 17:16 |
| fungi | clarkb: just nova boot and then... use nova console once it claims to be active? | 17:17 |
| clarkb | fungi: ya, also your key should be baked into the image so you can try ssh'ing in yourself | 17:17 |
| fungi | i've had plenty of luck with rackspace's oob console but my experience with nova console has been hit and miss | 17:17 |
| clarkb | fungi: but if nodepool can't ssh in that isn't likely to work | 17:17 |
| *** soren_ has quit IRC | 17:18 | |
| clarkb | fungi: hpcloud does also have a vnc console thing if you want interactive but we don't have logins | 17:18 |
| clarkb | nova console-log should work though | 17:18 |
| fungi | clarkb: it's "Timeout waiting for ssh access" but i don't recall if that actually means tcp timeout or could be a variety of errors hiding behind it | 17:18 |
| mordred | I build images with a root password in them when I'm debugging things like this | 17:18 |
| mordred | so that I can log in with the console | 17:18 |
| fungi | not a bad idea | 17:19 |
| *** mrunge has quit IRC | 17:19 | |
| clarkb | just don't do it via nodepool or our password will likely get logged :) | 17:19 |
| clarkb | s/our/your/ | 17:19 |
| fungi | heh, indeed | 17:19 |
| *** mpavone has quit IRC | 17:19 | |
| mordred | funzo_: 41 | 17:19 |
| mordred | echo "root:foobar" | chpasswd | 17:19 |
| clarkb | but my first guess is we need cloud-init for something during boot and we aren't getting that because cloud-init is uninstalled | 17:19 |
| mordred | fungi: I put that line into an install.d element | 17:20 |
| fungi | mordred: ooh, nice and simple | 17:20 |
| *** harlowja has quit IRC | 17:20 | |
| clarkb | which is odd because dhcp assigns the address, ssh host keys should generate by init if they don't exist, and we have the users baked into the image | 17:20 |
| *** harlowja has joined #openstack-infra | 17:20 | |
| clarkb | I suppose that maybe cloud-init is needed to configure the interface for dhcp and if so I will cry | 17:20 |
| mordred | can I suggest taht if we're needing to debug boot issues with new images we're making we do it on top of the minimal element where we're just gone through this debug cycle? | 17:20 |
| fungi | clarkb: yeah, entirely possible. we effectively can't install anything depending on python-setuptools because we uninstall it early and it can't be reinstalled via yum | 17:21 |
| clarkb | mordred: no bceause this is a snapshot image | 17:21 |
| clarkb | mordred: and it was working 1000 hours ago | 17:21 |
| mordred | 41 | 17:21 |
| mordred | gah. fungi said centos-6 dib image above | 17:21 |
| *** amuller has joined #openstack-infra | 17:21 | |
| fungi | right | 17:21 |
| *** shashankhegde has joined #openstack-infra | 17:21 | |
| clarkb | oh am I mixing the streams between images being fixed yesterday and other images being fixed yesterday? | 17:21 |
| fungi | clarkb and i are hunting separate centos 6.x image issues | 17:21 |
| clarkb | bare-centos6 were fixed yesterday | 17:21 |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Raise the cap on oslo.config to <1.7.0 https://review.openstack.org/182765 | 17:21 |
| clarkb | gah ok my bad | 17:22 |
| amuller | Neutron patches on the check queue are failing with 'substitution key 'envbindir' not found' | 17:22 |
| amuller | is this known already? | 17:22 |
| clarkb | amuller: can you provide a link to the failure? | 17:22 |
| amuller | clarkb: here's an example: https://jenkins06.openstack.org/job/gate-neutron-python27/5903/console | 17:22 |
| samueldmq | anteaya, hi - yes I had forgot that .. will add thanks :) | 17:22 |
| mordred | I mean, I know it's a slightly different stream - but if centos-6 dib is currently not working perhaps using centos-minimal and simple-init might be a path forward where the debugging involved gets us a win. otoh, if we don't want to do that, that's also totally understandable | 17:23 |
| clarkb | guess what released yesterday? tox | 17:23 |
| mordred | YAY! | 17:23 |
| mordred | clarkb: is that breaking new things yet? | 17:23 |
| fungi | mordred: maybe. i haven't actually looked at the centos element to see what it's doing | 17:23 |
| amuller | clarkb: I'll update tox locally, see if it happens | 17:23 |
| clarkb | mordred: yes I think it is the sourec of amuller's failure | 17:23 |
| clarkb | amuller: thanks | 17:23 |
| *** maishsk has joined #openstack-infra | 17:23 | |
| mordred | clarkb: yay! | 17:23 |
| fungi | woo! so tox broke parameter substitution in new and exciting ways this time? | 17:24 |
| mordred | 2015-05-13 16:40:41.482 | tox.ConfigError: ConfigError: substitution key 'envbindir' not found | 17:24 |
| mordred | that's so nice | 17:24 |
| *** claudiub has joined #openstack-infra | 17:24 | |
| amuller | mordred: clarkb: Updated tox locally and it's happening now as well | 17:24 |
| amuller | so what's the next move? | 17:24 |
| clarkb | amuller: either we update neutron (and other) tox.ini files to support new tox or we pin tox again | 17:25 |
| *** sarob has joined #openstack-infra | 17:25 | |
| fungi | https://bitbucket.org/hpk42/tox/issues | 17:25 |
| clarkb | amuller: I would like to see if we can make it work with current tox if possile | 17:25 |
| *** claudiub|2 has quit IRC | 17:26 | |
| *** bhuvan-afk is now known as bhuvan | 17:26 | |
| mordred | yay for major releases | 17:26 |
| clarkb | I'm not seeing anything related to envdir specifically in their changelog | 17:26 |
| mordred | it's a 2.0 though | 17:26 |
| clarkb | "introduce parser.add_testenv_attribute() to register an ini-variable for testenv sections." is possibly related | 17:26 |
| clarkb | mordred: yes | 17:26 |
| fungi | i don't see any bugs filed about it yet | 17:26 |
| clarkb | mordred: but they have never done semver before | 17:26 |
| *** dannywilson has quit IRC | 17:26 | |
| clarkb | mordred: so I don't necessarily treat that as special, instead they incremented .9 and got 1.0 | 17:27 |
| clarkb | mordred: add that to the 1.9 == 2.0 | 17:27 |
| clarkb | (sorry I am cranky about how they have done versions in the past0 | 17:27 |
| fungi | that's openbsd's version scheme | 17:27 |
| mordred | envbindir does not show up in the new --help-ini command | 17:27 |
| mordred | envdir does | 17:27 |
| mordred | so I suppose we could change {envbindir} to {envdir}/bin | 17:28 |
| *** bswartz has quit IRC | 17:28 | |
| clarkb | mordred: ++ | 17:28 |
| clarkb | amuller: ^ want to test that? | 17:28 |
| amuller | clarkb: trying | 17:28 |
| mordred | wsme is also affected | 17:29 |
| mordred | and networking-bigswitch and requests-mock and python-manilaclient | 17:29 |
| clarkb | the other thing that may bite us is they don't pass the env through by default | 17:29 |
| mordred | so 5 repos isn't that bad | 17:29 |
| fungi | http://tox.readthedocs.org/en/latest/config.html#substitutions-for-virtualenv-related-sections says it should exist in 2.0 | 17:29 |
| *** cdent has quit IRC | 17:30 | |
| *** salv-orlando has joined #openstack-infra | 17:30 | |
| clarkb | fungi: likely a bug in either docs or code then | 17:30 |
| clarkb | we have been good about setting the env via setenv though so don't expect many problems with not passing env through | 17:30 |
| fungi | yeah, should probably be filed. i can open one up with them | 17:30 |
| *** fawadkhaliq has joined #openstack-infra | 17:31 | |
| clarkb | so ya, if we can just make that small change for now and cruise along with 2.0.0 that would be best | 17:31 |
| clarkb | amuller: if you confirm that makes things happy, we/you/us can push patches to the other affected repos too | 17:33 |
| clarkb | and in the maen time I am going to delete some bare-trust images | 17:34 |
| clarkb | unless we want this to be self testing? | 17:34 |
| amuller | clarkb: yeah it works | 17:34 |
| amuller | clarkb: so I can push a patch for neutron. Should each project fix this now? How many are affected? | 17:34 |
| *** asselin has joined #openstack-infra | 17:35 | |
| clarkb | amuller: mordred has a list in scrollback, its not that many, neutron is probably the busiest | 17:36 |
| mordred | neutron, wsme, networking-bigswitch and requests-mock and python-manilaclient | 17:36 |
| *** marcusvrn1 has joined #openstack-infra | 17:36 | |
| *** mrmartin has joined #openstack-infra | 17:37 | |
| *** doude has quit IRC | 17:38 | |
| *** yamahata has quit IRC | 17:40 | |
| *** ivar-lazzaro has joined #openstack-infra | 17:40 | |
| *** reed_ has quit IRC | 17:41 | |
| *** marcusvrn has quit IRC | 17:41 | |
| *** marcusvrn_ is now known as marcusvrn | 17:41 | |
| *** alexpilotti has quit IRC | 17:41 | |
| openstackgerrit | Samuel de Medeiros Queiroz proposed openstack-infra/project-config: Experimental Devstack job for Identity v3 API only https://review.openstack.org/179661 | 17:42 |
| samueldmq | anteaya, ^ now I think it's ok | 17:42 |
| samueldmq | to go :) | 17:42 |
| clarkb | so ya I thinkwe should just update tox.ini for now, if we run into more problems we can think about pinning, but pinning tox results in unfun for devs because its hard for us to communicate the specific tox version to use | 17:42 |
| *** salv-orlando has quit IRC | 17:44 | |
| *** Hal1 has quit IRC | 17:44 | |
| *** ivar-lazzaro has quit IRC | 17:45 | |
| *** jtomasek has quit IRC | 17:45 | |
| anteaya | samueldmq: great thanks, I'll wait for jenkins to report on it | 17:46 |
| *** dguerri is now known as _dguerri | 17:46 | |
| clarkb | mordred: shade 0.6.1 is on pypi now. Ready to approve the use shade in nodepool chnage? | 17:47 |
| samueldmq | anteaya, fair enough :) | 17:48 |
| mordred | clarkb: you don't still have yoru nodepool/devstack aroound do you? | 17:49 |
| mordred | clarkb: we did make a change to os-client-config since you tested and I'm waffling on whether I think we should care | 17:49 |
| clarkb | mordred: I do not, but its easy enough to hold a neutron test node and set nodepool up when its done | 17:49 |
| clarkb | ah | 17:49 |
| mordred | I mean, it's tested and all and has passed all the tests - but still - you know how it is | 17:49 |
| anteaya | samueldmq: thanks | 17:49 |
| *** patrickeast has joined #openstack-infra | 17:50 | |
| openstackgerrit | Ruby Loo proposed openstack/requirements: Use httpretty < 0.8.7 to avoid broken unittests https://review.openstack.org/182340 | 17:50 |
| *** yfried__ has joined #openstack-infra | 17:51 | |
| clarkb | mordred: what changed in os-c-c? | 17:56 |
| *** yamamoto has quit IRC | 17:57 | |
| *** yamamoto has joined #openstack-infra | 17:57 | |
| clarkb | mordred: I just held a rax node running neutron-full can set up nodepool there shortly | 17:57 |
| *** yamamoto has quit IRC | 17:58 | |
| mordred | clarkb: two things. the change to handle defaults processing in a slightly different order, and the use of appdirs library rather tahn hardcoded XDG_CONFIG env vars | 17:58 |
| mordred | clarkb: it SHOULD be completely no op to shade/nodepool | 17:58 |
| mordred | it SHOULD be no op to everything, honestly | 17:58 |
| *** fawadk has joined #openstack-infra | 17:58 | |
| *** marcusvrn has quit IRC | 17:59 | |
| mordred | but I'm paranoid because this is the first shade-into-infra-production patch :) | 17:59 |
| clarkb | ya np, its a quick setup to check now that I have the whole thing written down | 18:00 |
| *** e0ne is now known as e0ne_ | 18:00 | |
| clarkb | then its just waiting for qemu to do its thing | 18:00 |
| sdague | zaro: so I don't know, I threw out a post on the gerrit ML with as much detail as I could, in the hopes that it would create pointers - https://groups.google.com/forum/#!topic/repo-discuss/MYiSeRG5Ras | 18:00 |
| mordred | clarkb: this is https://review.openstack.org/#/c/181645/ the patch I'm worried about - since there _is_ a behavior change in that patch | 18:01 |
| sdague | it seems to have gathered one response so far | 18:01 |
| *** nfedotov has joined #openstack-infra | 18:01 | |
| *** fawadkhaliq has quit IRC | 18:01 | |
| *** tiswanso has quit IRC | 18:01 | |
| clarkb | greghaynes: when you tried on hpcloud what flavor did you use? I need to just give it a go on our hpcloud nodes to see if it works at all. If it does next step is a test | 18:01 |
| fungi | okay, getting the infra docs move back on track, openstackid and storyboard haven't published to http://docs.openstack.org/infra/http://docs.openstack.org/infra/<reponame>/ yet. not sure if they haven't merged new patches in a while, or if their publication jobs are otherways broken | 18:02 |
| *** yamamoto has joined #openstack-infra | 18:02 | |
| fungi | er, to http://docs.openstack.org/infra/<reponame>/ | 18:02 |
| * fungi needs to replace his paste button | 18:02 | |
| sdague | zaro: I got as far as working step debugger in eclipse on the gerrit code, but I have to say as I'm watching the relevant section of gerrit here, I'm lost pretty quickly in their data structures | 18:02 |
| *** yamamoto has quit IRC | 18:02 | |
| mordred | fungi: I think storyboard has released patches | 18:02 |
| jeblair | sdague: wow. would you like to take over communication with the gerrit community regarding major data corruption bugs? | 18:02 |
| krotscheck | fungi: We landed one monday morning | 18:03 |
| zaro | sdague: maybe i should restore to allow others to maybe pickup? | 18:03 |
| *** Sukhdev_ has joined #openstack-infra | 18:03 | |
| greghaynes | clarkb: using to 30something gb ram | 18:03 |
| jeblair | fungi: i'll dig into openstackid | 18:03 |
| *** annegentle has quit IRC | 18:04 | |
| sdague | jeblair: honestly, I think with the other things on my plate you probably don't want me in the critical path for that. | 18:04 |
| *** tiswanso_ has joined #openstack-infra | 18:04 | |
| fungi | jeblair: thanks--i'm checking the logs for storyboard | 18:04 |
| *** fawadkhaliq has joined #openstack-infra | 18:04 | |
| fungi | krotscheck: thanks | 18:04 |
| krotscheck | fungi: No prob | 18:04 |
| zaro | sdague: yeah, that's probably the most magical part of gerrit i've seen so far. i can't figure out how it works with lucene. | 18:04 |
| *** sergsh has quit IRC | 18:04 | |
| *** e0ne_ is now known as e0ne | 18:04 | |
| mordred | sdague: well, I think it's that we sent a data corruption bug and got no responses but got one response, so so far you are infinitely better :) | 18:05 |
| greghaynes | mordred: my similar trick for debugging is to add the 'devuser' element and set the password | 18:05 |
| greghaynes | mordred: re: setting a root pass | 18:05 |
| mordred | greghaynes: ++ | 18:05 |
| greghaynes | (that is precisely why I maded that element) | 18:05 |
| *** Sukhdev has quit IRC | 18:05 | |
| sdague | zaro: it's building up some kind of predicate stack, but at some point .... but I think as was listed in the response at some point it's just dropped in an optimization | 18:06 |
| *** yamahata has joined #openstack-infra | 18:06 | |
| *** e0ne has quit IRC | 18:06 | |
| clarkb | greghaynes: thanks, thats what we use in thegate too :/ | 18:06 |
| *** fawadk has quit IRC | 18:06 | |
| fungi | mordred: i think maybe we missed switching the docs job for storyboard http://logs.openstack.org/b0/b0e82d23cd170f0b5d2e987e8479b9534c3d5c3e/post/storyboard-docs/df2c9e4/console.html | 18:06 |
| fungi | patching that momentarily | 18:06 |
| mordred | fungi: oops! | 18:07 |
| * clarkb waits for lxml to build | 18:07 | |
| mordred | clarkb: it's your favorite thing | 18:09 |
| mordred | clarkb: you know you could have just apt-get installed python-lxml ... | 18:09 |
| clarkb | ya :( | 18:09 |
| dstufft | wheel you some lxml | 18:09 |
| sdague | jeblair: so there is also this thread on the forum - https://groups.google.com/forum/#!searchin/repo-discuss/PackFile/repo-discuss/o-HPPbwYn7A/UWcqe7zSXIoJ - which looks very similar | 18:10 |
| *** ajo has quit IRC | 18:10 | |
| *** fawadk has joined #openstack-infra | 18:11 | |
| greghaynes | clarkb: oh, you mentioned a while ago that we have an openstack clouds.yaml getting deployed onto the nodepool node? | 18:11 |
| sdague | jeblair: what version of git is on that machine? | 18:12 |
| clarkb | greghaynes: ya I think that merged let me double check | 18:12 |
| jeblair | sdague: yeah, we _also_ saw that, but we read that thread and concluded that we could solve that fairly easily by changing how we repack; at any rate, that only happens on sunday, so that was a one time thing. | 18:12 |
| sdague | ok | 18:12 |
| greghaynes | clarkb: ok, we should make sure that lines up with the nodepool.yaml providers and isnt going to override anything... | 18:12 |
| clarkb | greghaynes: it shouldn't beacuse we don't have a cloud right? | 18:12 |
| greghaynes | yep | 18:13 |
| clarkb | greghaynes: sorry, we don't specify cloud: foo in nodepool.yaml yet | 18:13 |
| greghaynes | ok, yea | 18:13 |
| greghaynes | so should be fine | 18:13 |
| mordred | yah. we start by just passing in args to the OpenStackCloud constructor | 18:13 |
| *** fawadkhaliq has quit IRC | 18:13 | |
| clarkb | ok so that should be fine | 18:14 |
| *** fawadk has quit IRC | 18:14 | |
| *** megm_ has quit IRC | 18:15 | |
| *** amuller is now known as amuller_afk | 18:16 | |
| *** gokrokve has joined #openstack-infra | 18:16 | |
| *** megm has joined #openstack-infra | 18:16 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/requirements: [WIP] Remove dev-requirements.txt https://review.openstack.org/182776 | 18:17 |
| *** maishsk_ has joined #openstack-infra | 18:18 | |
| *** angela-s has joined #openstack-infra | 18:18 | |
| *** maishsk has quit IRC | 18:18 | |
| *** maishsk_ is now known as maishsk | 18:18 | |
| *** _david_ has joined #openstack-infra | 18:19 | |
| *** nfedotov has quit IRC | 18:19 | |
| jeblair | fungi, mordred: openstackid has not merged a change since the publish job was changed | 18:19 |
| jeblair | (last was the day before) | 18:19 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/project-config: Publish storyboard docs to docs.o.o https://review.openstack.org/182778 | 18:19 |
| fungi | jeblair: perfect. that's easy | 18:20 |
| *** nfedotov has joined #openstack-infra | 18:20 | |
| fungi | and 182778 hopefully solves storyboard | 18:20 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Publish traning-guide files from master again https://review.openstack.org/182779 | 18:20 |
| openstackgerrit | Boris Pavlovic proposed openstack-infra/project-config: Add rally job to nova experimental queue https://review.openstack.org/147765 | 18:20 |
| jeblair | there's a pending change, but it's 3 kilolines of diff, so i'm not going to review it. | 18:20 |
| fungi | jeblair: also, php :/ | 18:20 |
| fungi | i'll go ahead and test out the redirects too since they should in theory be in place just don't have dns pointed at them quite yet | 18:21 |
| jeblair | i mean, it's mostly removing stuff that probably shouldn't be there, like the foundation marketplace | 18:21 |
| pleia2 | mordred (or anyone else, but he +2), dependency merged, any other reason we should wait on this one? https://review.openstack.org/#/c/180624/ | 18:21 |
| fungi | yep, that's moved to the openstackid-resources repo now | 18:21 |
| jeblair | fungi: so i'll work on a manual jenkins trigger of openstackid master | 18:22 |
| fungi | jeblair: thanks. i'll do the same for storyboard once the jobs for it update, unless there's a lhf patch i can review and approve there at that point | 18:22 |
| *** rlandy has joined #openstack-infra | 18:24 | |
| fungi | i thought i had manually tested out the redirects, but it looks like the system-config redirect is getting in the way of the others. i'll work out a fix real fast | 18:24 |
| *** ddieterly has quit IRC | 18:25 | |
| harlowja | periodic poke for someone to review https://review.openstack.org/#/c/171393/ :) | 18:25 |
| harlowja | thx | 18:25 |
| fungi | oh... this might be a bit of a problem. having a permanent redirect at ^/(.*) means that browsers are going to root everything there... i think i actually need explicit redirects for each current file in system-config instead | 18:26 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Publish traning-guide files from master again https://review.openstack.org/182779 | 18:26 |
| fungi | so as to avoid stepping on the other sub-sites | 18:27 |
| fungi | hrm, or maybe this can be done with exclusions | 18:27 |
| *** amitgandhinz has joined #openstack-infra | 18:27 | |
| mordred | clarkb: I've got to step away for a few - if you're comfortable that you've tested this version, I'm fine with you approving - or you can wait til I'm back | 18:27 |
| * greghaynes has to go to lunch for a bit | 18:28 | |
| clarkb | mordred: kk, just about to turn nodepool on in the test node | 18:28 |
| greghaynes | so might be a good idea to wait until after | 18:28 |
| clarkb | mordred: it will probably take about 20 minutes for it to spit out results | 18:28 |
| greghaynes | or not, up to you ;) | 18:28 |
| jeblair | https://jenkins01.openstack.org/job/openstackid-infra-docs/1/console | 18:29 |
| *** otter768 has joined #openstack-infra | 18:29 | |
| *** tonytan4ever has joined #openstack-infra | 18:29 | |
| AJaeger | jeblair: do you have any idea why I get on https://review.openstack.org/182779 "This change depends on a change that failed to merge." ? | 18:29 |
| openstackgerrit | Joshua Harlow proposed openstack-infra/project-config: Add non-voting pypy jobs to taskflow https://review.openstack.org/182782 | 18:30 |
| gordc | fungi: i think we need a pycadf stable/icehouse | 18:30 |
| AJaeger | I initially pushed 182779 first and then the dependend one which is https://review.openstack.org/#/c/182780/ . Is that higher number a problem? | 18:31 |
| gordc | http://logs.openstack.org/30/154530/2/gate/gate-grenade-dsvm/9cd01f0/logs/grenade.sh.txt.gz#_2015-05-13_17_41_29_232 | 18:31 |
| AJaeger | jeblair: this is using "Depends-On" | 18:31 |
| jeblair | AJaeger: looking | 18:31 |
| gordc | fungi: it looks like new pycadf needs pytz<=2014.10 but pycadf 0.5.1 (which icehouse uses) has pytz>=2010h | 18:32 |
| AJaeger | thanks, jeblair ! | 18:32 |
| jeblair | fungi: openstackid is fixed | 18:32 |
| fungi | jeblair: thanks! | 18:33 |
| * fungi is still fiddling with mod_rewrite | 18:33 | |
| *** otter768 has quit IRC | 18:33 | |
| *** spzala has joined #openstack-infra | 18:33 | |
| AJaeger | jeblair: same error message with https://review.openstack.org/#/c/182773/ | 18:33 |
| *** yfried has joined #openstack-infra | 18:34 | |
| *** claudiub|2 has joined #openstack-infra | 18:34 | |
| *** claudiub has quit IRC | 18:34 | |
| _david_ | zaro, Around? I can not reproduce https://code.google.com/p/gerrit/issues/detail?id=3365 | 18:35 |
| lifeless | mordred: 182438 needs to land urgently | 18:36 |
| lifeless | mordred: if it hasn't already | 18:36 |
| *** yfried__ has quit IRC | 18:36 | |
| lifeless | ah, it has. grand. | 18:36 |
| jeblair | AJaeger: i think there's a bug where if you upload a new patchset of a dependent change, it ejects the change(s) that need(s) it | 18:36 |
| zaro | _david_: it's here: https://review-dev.openstack.org/#/dashboard/self | 18:36 |
| *** jlanoux_ has joined #openstack-infra | 18:36 | |
| *** jlanoux has quit IRC | 18:37 | |
| _david_ | zaro /me don't have access to this machine | 18:37 |
| zaro | _david_: anonymous access? | 18:37 |
| jeblair | AJaeger: it's actually not straightforward to fix; i'm not certain what the right thing to do in that situation is; i suspec it would be to restart jobs based on the new patchset of the dependent change. that will require a bit of development to implement. | 18:37 |
| AJaeger | jeblair: is there anything I can do? Wait for jenkins to finish building? | 18:38 |
| zaro | _david_: https://review-dev.openstack.org/ | 18:38 |
| _david_ | zaro Ah, now I see it: https://review-dev.openstack.org//#/dashboard/self | 18:38 |
| jeblair | AJaeger: yeeah, your PS2 should be running now with the latest/correct PS of the dependent change, so all should be good soon. | 18:38 |
| AJaeger | jeblair: ok, then I wait. Thanks | 18:38 |
| jeblair | AJaeger: (zuul is doing the right thing, just not the convenient thing; we can make it smarter) | 18:39 |
| AJaeger | ;) | 18:39 |
| _david_ | zaro I tried to reproduce it locally, in my environment (on gerrit master) and I can't. Why? | 18:39 |
| zaro | _david_: actually i see you've made changes recently to openid so thought you mind know something but upon further inspection of git history i think this is the change that caused the problem. https://gerrit.googlesource.com/gerrit/+/71891446a5f05fd9b015236144c206ac70043f21 | 18:40 |
| *** arxcruz has quit IRC | 18:41 | |
| _david_ | zaro, None of my recent OpenID changes were merged prior to 2.10.3, and you are using 2.10.2 | 18:42 |
| _david_ | zaro, but still i cannot reproduce it. Why? | 18:43 |
| *** e0ne has joined #openstack-infra | 18:44 | |
| _david_ | zaro my settings are: canonicalWebUrl = http://localhost:8080, type = OPENID_SSO, openIdSsoUrl = https://login.launchpad.net/+openid and all is fine after login | 18:44 |
| zaro | _david_: ohh maybe because we are using from our fork: http://git.openstack.org/cgit/openstack-infra/gerrit/log/?h=openstack/2.10.2 | 18:44 |
| _david_ | (on most recent master) | 18:44 |
| _david_ | zaro, What is simpler: you try vanilla master or /me try your fork? ;-) | 18:45 |
| jeblair | _david_, zaro: i believe our 2.10.2.23 is equivalent to 2.10.3 from what zaro tells me about the additional patches onto our branch | 18:45 |
| jeblair | (and it's a very shallow fork) | 18:45 |
| _david_ | jeblair, Do you have my JGit fix ? | 18:45 |
| jeblair | _david_: http://git.openstack.org/cgit/openstack-infra/gerrit/log/?h=openstack/2.10.2 | 18:46 |
| *** claudiub has joined #openstack-infra | 18:46 | |
| zaro | _david_: our fork has jgit ver 3.7.0.201502260915-r.58-g65c379e | 18:47 |
| _david_ | zaro, Oh NO. This version is horribly broken | 18:48 |
| jeblair | there seem to be a lot of changes in .3 that are not in our 2.10 branch | 18:48 |
| _david_ | jeblair, 2.10.3 and 2.10.3.1 are broken, because they don't have my JGit fix | 18:49 |
| *** claudiub|2 has quit IRC | 18:50 | |
| fungi | _david_: we dug through the jgit 4 changelog and didn't see any obvious changes which would have fixed the behavior we were getting | 18:51 |
| _david_ | jeblair, zaro https://code.google.com/p/gerrit/issues/detail?id=3300 | 18:51 |
| *** Sukhdev_ has quit IRC | 18:51 | |
| _david_ | fungi, different story. We had severe performance degradation on both git push and replication. | 18:52 |
| clarkb | performance degratdation isn't the problem | 18:52 |
| fungi | _david_: yep, did see that discussion | 18:52 |
| jeblair | ah, yeah; i don't think we saw that, though maybe we would have eventually if gerrit hadn't thrown out the largest packfiles in our repos :) | 18:52 |
| jeblair | (it's hard to be slow if you aren't actually doing anything) | 18:52 |
| *** yapeng has joined #openstack-infra | 18:54 | |
| _david_ | jeblair, Have you tried to revert the change you were referencing in your post on repo-discuss ? | 18:54 |
| *** banix has quit IRC | 18:54 | |
| _david_ | jeblair, and build custom version of JGit and try it if it solves the problem? | 18:55 |
| clarkb | greghaynes: my hpcloud test node built an image, checking if it can boot it now | 18:56 |
| fungi | _david_: unfortunately that change was implemented in 3.5, so could be a lot to untangle between there and whatever gerrit 2.10 needs, but no i don't believe we tried to do that | 18:56 |
| *** baoli has quit IRC | 18:57 | |
| jeblair | _david_: we did try gerrit 2.10 with jgit 3.2, but that didn't work for other reasons. | 18:57 |
| fungi | _david_: so far we don't have the ability to reproduce what we saw without another lengthy production outage. we're working on figuring out a better way to make our dev server mirror our production content and activity | 18:57 |
| fungi | in hopes that we can reproduce the failure on demand | 18:57 |
| _david_ | fungi, jeblair It would be indeed very helpful if you guys could reproduce it. | 18:58 |
| jeblair | (also, i don't actually think there's an error in that change; or if there is, we haven't found it yet) | 18:58 |
| fungi | _david_: we agree | 18:58 |
| _david_ | fungi, jeblair I have some connections to JGit maintainers, here at SAP they could and would help, ... | 18:59 |
| fungi | would open us up to all sorts of investigation when we're not preventing our developers from gettign work done | 18:59 |
| *** _dguerri is now known as dguerri | 18:59 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/system-config: Correct ci.o.o redirects so slashes are optional https://review.openstack.org/182790 | 18:59 |
| fungi | jeblair: ^ turns out the redirects were *mostly* working | 18:59 |
| *** annegentle has joined #openstack-infra | 19:00 | |
| jeblair | _david_: cool, thanks -- we'll ping you when we're able to reproduce; and if you have any more ideas of where to look in the mean time, let us know | 19:00 |
| clarkb | greghaynes: yup node just went read so not sure what issues you were seeing | 19:00 |
| clarkb | greghaynes: but I think this means i can make a test | 19:00 |
| *** Longgeek has quit IRC | 19:01 | |
| *** Longgeek has joined #openstack-infra | 19:01 | |
| _david_ | jeblair, I have one idea: stop using broken JGit version. | 19:01 |
| *** Svedrin has quit IRC | 19:01 | |
| *** Longgeek has quit IRC | 19:02 | |
| jeblair | _david_: that is a good one, probably first thing we should test when we repro. ;) | 19:02 |
| *** Longgeek has joined #openstack-infra | 19:02 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/project-config: Publish storyboard docs to docs.o.o https://review.openstack.org/182778 | 19:02 |
| fungi | missed a required parameter ^ | 19:02 |
| _david_ | jeblair, https://git.eclipse.org/r/31081 was very serious regression | 19:02 |
| *** yapeng has quit IRC | 19:02 | |
| *** claudiub|2 has joined #openstack-infra | 19:03 | |
| *** soren_ has joined #openstack-infra | 19:03 | |
| *** HeOS has joined #openstack-infra | 19:03 | |
| jeblair | i have to run to lunch; biab | 19:03 |
| *** rockyg has joined #openstack-infra | 19:03 | |
| *** Longgeek has quit IRC | 19:04 | |
| *** carl_baldwin has quit IRC | 19:04 | |
| *** Longgeek has joined #openstack-infra | 19:04 | |
| _david_ | jeblair, I happen to know, that Google immediately reverted it for all their products. So seriously stop using JGit with this commit inside and move to either: 3.7.1.201504261725-r or 4.0.0.201505050340-m2 | 19:04 |
| AJaeger | fungi, that means we can publish http://docs.openstack.org/infra/ once your patch is in. Great! | 19:05 |
| zaro | _david_: we've already reverted to 2.8 with jgit ver 3.2 | 19:05 |
| *** claudiub has quit IRC | 19:05 | |
| AJaeger | (and another storyboard patch landed) | 19:05 |
| AJaeger | fungi: thanks for getting all links working! | 19:05 |
| krotscheck | Eh? Wha? | 19:05 |
| *** sdake has joined #openstack-infra | 19:06 | |
| *** ivar-lazzaro has joined #openstack-infra | 19:06 | |
| AJaeger | krotscheck: http://docs.openstack.org/infra/ references storyboard but clicking on it gives a 404 | 19:06 |
| _david_ | zaro, Ah I see, only https://review-dev.openstack.org/ on 2.10 | 19:06 |
| zaro | _david_: you used the latest master to repro the extra forward slash? | 19:07 |
| krotscheck | AJaeger: Ah, yes. | 19:07 |
| _david_ | zaro, Yes, I did. | 19:07 |
| krotscheck | I take it we want a storyboard patch to land so that docs can become a thing? | 19:07 |
| fungi | AJaeger: jeblair took care of the missing openstackid, 182778 solves storyboard (once it lands and we either merge a random storyboard change or manually trigger the post job), then 182790 would be a good idea to make the redirects from teh old urls more robust before i adjust dns | 19:07 |
| _david_ | zaro just curious, how you downgraded gerrit from 2.10 => 2.8? | 19:07 |
| zaro | _david_: i would expect it to repro there as well. well i can give that a try and update the issue. | 19:07 |
| krotscheck | Ok, lemme see if I have anything useful to land. | 19:08 |
| * krotscheck doesn't think so. | 19:08 | |
| zaro | _david_: we went throught the changes and reverted on db. | 19:08 |
| AJaeger | fungi: once storyboard is up, we can merge https://review.openstack.org/#/c/176921/ as well... | 19:08 |
| fungi | _david_: mordred write up the necessary queries to reverse the db migrations | 19:08 |
| _david_ | zaro Wau, respect. I'm asking because some folks were really interested in Gerrit downgrade feature | 19:09 |
| AJaeger | jeblair: thanks for the openstackid publishing! | 19:09 |
| zaro | _david_: https://review.openstack.org/#/c/182113/1/tools/gerrit-rollback-2.10-2.8.sql | 19:09 |
| fungi | _david_: it's nontrivial because of how some parts of the migrations are inferred from changes in the java source | 19:09 |
| krotscheck | fungi: I've got something. | 19:09 |
| _david_ | fungi, Hey, cool, any chance he can contribute it back to the community? | 19:09 |
| krotscheck | fungi: Ready to land? | 19:09 |
| AJaeger | krotscheck: not yet | 19:09 |
| krotscheck | AJaeger: KK, ping me when you're ready. | 19:09 |
| AJaeger | first 182778 needs to be merged and in jenkins | 19:10 |
| fungi | _david_: and we identified at least one incompatible account preference someone changed in the 2 days we were running 2.10 which had to be manually filtered after the fact to get their access working again | 19:10 |
| AJaeger | fungi, could you ping krotscheck, please? I won't be around that long... | 19:10 |
| _david_ | fungi, *LOL* | 19:10 |
| fungi | AJaeger: yep | 19:11 |
| openstackgerrit | Yanis Guenane proposed openstack-infra/tripleo-ci: Specify Redis Virtual IP https://review.openstack.org/182795 | 19:11 |
| fungi | _david_: it was far from a clean process, but our other options were exhausted and the dangers in the rollback cleanup were outweighed by having our code review system basically broken any longer than we already had | 19:12 |
| *** Rockyg_ has joined #openstack-infra | 19:12 | |
| *** notnownikki has quit IRC | 19:12 | |
| annegentle | flaper87: hey, which timezone are you in? Ping if you're awake :) | 19:12 |
| _david_ | fungi, Sure. I just wondering, that you did upgrade without healthy checks on https://review-dev.openstack.org/ on 2.10 | 19:13 |
| *** panda has quit IRC | 19:13 | |
| fungi | _david_: we tested that same build of 2.10 rigorously on review-dev | 19:13 |
| _david_ | fungi, But not under heavy load? | 19:14 |
| fungi | but clearly the volume and character of our production data and interactions was not at all well modeled | 19:14 |
| *** panda has joined #openstack-infra | 19:14 | |
| fungi | _david_: right, we tested for broken features but not for regressions under load | 19:14 |
| zaro | but we are working to fix that going forward | 19:14 |
| fungi | yep | 19:14 |
| *** rockyg has quit IRC | 19:15 | |
| _david_ | fungi, zaro, interestingly, we are running for 1 month on 2.10.x and havn't seen any issues. We have big repos too 2GB+ but apparently not the load you guys have. | 19:16 |
| fungi | _david_: around 2k active developers and >600 git repositories | 19:16 |
| *** Sukhdev has joined #openstack-infra | 19:16 | |
| sdague | fungi: + 200 ci systems watching the stream and reporting changes back | 19:17 |
| fungi | yes, that | 19:17 |
| _david_ | fungi, Just for th record, SAP have 10K devs, and 4000 projects. And i happen to know, that they are running on 2.10.2 without any issue | 19:17 |
| *** Svedrin has joined #openstack-infra | 19:17 | |
| _david_ | sdague, OK | 19:17 |
| sdague | _david_: what's your turnover time on 10K new revisions? | 19:17 |
| fungi | _david_: good to know. that makes me feel a little better. are they using filesystem backend? we've heard rumors that most of the large gerrit installations store their git content in databases | 19:18 |
| sdague | we're doing that about every 40 days | 19:18 |
| _david_ | sdague, I don't know, I'm not affiliated, just heard from the discussion from SAP guys | 19:18 |
| sdague | so, I looked, and that's about +50% over what android's gerrit is doing on rate | 19:19 |
| sdague | and we do a lot more ps per rev (some revs have > 80 ps) | 19:19 |
| _david_ | sdague, Really? 80? respect, my record for inline edit ~50 ;-) | 19:20 |
| sdague | _david_: I am sure I can find something over 100 :) | 19:20 |
| _david_ | fungi, AFAICT, only Google is using GFS, all other folks use normal file system | 19:21 |
| _david_ | sdague, show me? ,-) | 19:21 |
| fungi | got it. thanks! | 19:21 |
| _david_ | fungi, My proposal: pick sane JGit version: 4.0.0.201505050340-m2, Or just pick 2.10.4. This version already contained it. | 19:22 |
| _david_ | fungi, And try hard to reproduce the breakage. One you are able to do it, it would be much simple to track it down. | 19:23 |
| sdague | _david_: man, if gerrit let me search by ps #, that would be awesome to go find them. :) | 19:23 |
| sdague | I'll see if I can hack up a search script | 19:23 |
| _david_ | sdague, MySQL query? | 19:24 |
| sdague | I don't have access to mysql | 19:24 |
| _david_ | sdague, You are Gerrit admin, aren't you? | 19:24 |
| sdague | fungi could | 19:24 |
| fungi | i have a feeling sdague doesn't want to be a gerrit admin. he already does rather a lot ;) | 19:24 |
| sdague | _david_: nope, just an enthusiast user | 19:24 |
| _david_ | sdague, Even not on https://review-dev.openstack.org/ ? | 19:24 |
| *** claudiub has joined #openstack-infra | 19:25 | |
| sdague | _david_: nope | 19:25 |
| sdague | _david_: if you have a moment though, I did post about the issue with group= not working in search on repo-discuss, any thoughts you have on how to fix that would be highly appreciated | 19:27 |
| _david_ | sdague, Yes i saw that, ... | 19:27 |
| *** claudiub|2 has quit IRC | 19:27 | |
| sdague | ok, got to do some family things, see everyone tomorrow | 19:28 |
| _david_ | sdague, I'm not familiar with this part of code, unfortunately and that's tough topic, because this feature hacked by Qualcomm guys, .. | 19:28 |
| fungi | database says our highest patchset count (236) on a change is in ci-sandbox, so that doesn't really count | 19:29 |
| *** pcrews has quit IRC | 19:29 | |
| zaro | _david_: i would continue to try working on that if i could get some help from anybody. | 19:30 |
| _david_ | fungi, I am loking for a change with max(ps) fabricated by human? | 19:30 |
| _david_ | looking | 19:30 |
| fungi | glance and swift are tied for the #2 spot at 162 patchsets but those are both by bots: https://review.openstack.org/27091 and https://review.openstack.org/88736 | 19:31 |
| _david_ | zaro, We learned very hard one thing contributing to gerrit: Don't give up | 19:31 |
| *** Swami_ has joined #openstack-infra | 19:31 | |
| fungi | aha! #4 place is held by this glance change with 148 patchsets, not made by a bot: https://review.openstack.org/136629 | 19:32 |
| zaro | _david_: haha! it was my first time to give up :) | 19:32 |
| *** e0ne has quit IRC | 19:32 | |
| *** Mmike has quit IRC | 19:32 | |
| bknudson | +2227 is going to be difficult to review | 19:33 |
| _david_ | fungi, unbelievable | 19:33 |
| fungi | roughly skimming the top patchset counts, we have about 10 at or over 100 made by human hands | 19:33 |
| *** e0ne has joined #openstack-infra | 19:33 | |
| mtreinish | fungi: heh, 148 and still going | 19:33 |
| fungi | mtreinish: right, it could gain sentience. give it time | 19:34 |
| zaro | ahh yeah, that's seems like the same as dead. | 19:34 |
| *** Swami has quit IRC | 19:35 | |
| clarkb | mordred greghaynes I am lunching but will shade nodepool in about half an hour | 19:35 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone service resource methods https://review.openstack.org/177621 | 19:36 |
| *** Mmike has joined #openstack-infra | 19:36 | |
| *** frickler has quit IRC | 19:36 | |
| *** amuller_afk is now known as amuller | 19:36 | |
| *** Mmike has quit IRC | 19:36 | |
| *** Mmike has joined #openstack-infra | 19:36 | |
| *** eharney has quit IRC | 19:38 | |
| *** enikanorov2 has joined #openstack-infra | 19:39 | |
| *** jlanoux_ has quit IRC | 19:40 | |
| *** packet has joined #openstack-infra | 19:40 | |
| notmyname | is it possible to do a "related-spec" to auto link to specs like we can with bugs and blueprints? | 19:41 |
| notmyname | in commit messages in gerrit | 19:41 |
| *** Mmike has quit IRC | 19:42 | |
| mtreinish | notmyname: isn't that what the bp is used for to do the tracking of the work item | 19:42 |
| SpamapS | _david_: BUT, have you seen how magnificant our rainbow bikeshed is? ;-) | 19:42 |
| fungi | notmyname: http://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/manifests/gerrit.pp#n142 | 19:42 |
| lifeless | notmyname: in principle yes but needs to be configured in gerrit | 19:43 |
| notmyname | fungi: thanks | 19:43 |
| lifeless | notmyname: I'm not sure if regexes can be scoped to projects | 19:43 |
| *** e0ne has quit IRC | 19:43 | |
| _david_ | SpamapS, "rainbow bikeshed" ? I'm afraid i missed that? | 19:44 |
| *** nfedotov1 has joined #openstack-infra | 19:44 | |
| *** stevemar has quit IRC | 19:44 | |
| *** Krinkle is now known as Krinkle|detached | 19:44 | |
| *** nfedotov has quit IRC | 19:44 | |
| fungi | lifeless: notmyname: yeah, it would probably need to be a subcomponent of a specs.o.o url | 19:44 |
| openstackgerrit | Steven Hardy proposed openstack/requirements: Increase stable/kilo python-heatclient upper cap to 0.6 https://review.openstack.org/182672 | 19:45 |
| *** banix has joined #openstack-infra | 19:45 | |
| *** turul_ has quit IRC | 19:45 | |
| fungi | lifeless: notmyname: given that it would also be spec lifecycle and project specific, probably specs.o.o needs some sort of normalized query interface to support that | 19:45 |
| fungi | otherwise you're stuck doing something like related-spec: openstack/swift-specs/specs/in_progress/at_rest_encryption | 19:46 |
| *** turul_ has joined #openstack-infra | 19:46 | |
| fungi | which quickly goes stale once it goes from in_progress to implemented | 19:46 |
| lifeless | its almost like you need a web site | 19:46 |
| lifeless | like blueprints.launchpad.net | 19:46 |
| notmyname | those regexes remind me of something that happened recently. someone referenced a Red Hat bug in the commit message (publicly viewable bug, so no issues). but gerrit auto-linked it to launchpad. which is just funny | 19:46 |
| *** teran has quit IRC | 19:47 | |
| fungi | notmyname: yeah, that supybot plugin assumes "bug somenumber" means lp, but it still recognizes redhat bugzilla entries by url if mentioned | 19:47 |
| mtreinish | lifeless: yes, that was my earlier point. I thought specs were just to review and define a work item. But the actually lifecycle of the work was still in lp | 19:47 |
| *** Sukhdev_ has joined #openstack-infra | 19:47 | |
| notmyname | fungi: I think this was something like "Red Hat bug: 1234" or something similar. which would match | 19:48 |
| openstack | bug 1234 in Launchpad itself "Gina is an unmaintainable mess of command line options, environment variables and shell scripts" [Medium,Fix released] https://launchpad.net/bugs/1234 - Assigned to Daniel Henrique Debonzi (debonzi) | 19:48 |
| *** Mmike has joined #openstack-infra | 19:48 | |
| *** Mmike has joined #openstack-infra | 19:48 | |
| *** Sukhdev has quit IRC | 19:50 | |
| SpamapS | _david_: are you not familiar with the term "bikeshedding" ? | 19:50 |
| *** yamamoto has joined #openstack-infra | 19:51 | |
| *** erlon has quit IRC | 19:51 | |
| *** hashar has joined #openstack-infra | 19:52 | |
| openstackgerrit | Merged openstack-infra/project-config: Publish storyboard docs to docs.o.o https://review.openstack.org/182778 | 19:53 |
| lifeless | SpamapS: technicolour rainbow bikeshed pleeease | 19:53 |
| greghaynes | clarkb: ok, i just finished lunch so ill be around | 19:53 |
| SpamapS | lifeless: I prefer chromakey rainbow bikeshed | 19:54 |
| fungi | the irony is, the house i'm buying does actually have a bike shed, and it's in need of a paint job. i'm thinking of doing a poll to find out what color i should paint it | 19:54 |
| SpamapS | fungi: nooooooooooooooooo | 19:54 |
| anteaya | fungi: ha ha ha | 19:54 |
| fungi | might make for an interesting photo essay | 19:55 |
| mtreinish | fungi: I vote for my normal etherpad color | 19:55 |
| mtreinish | you might be able to see it from space then | 19:55 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard: Single event email worker. https://review.openstack.org/179542 | 19:55 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard: Factory subject is now a template file. https://review.openstack.org/181763 | 19:55 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard: Revert "Email Outbox" https://review.openstack.org/182043 | 19:55 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone service resource methods https://review.openstack.org/177621 | 19:55 |
| anteaya | mtreinish: you are consistent with that colour | 19:55 |
| AJaeger | fungi: you could make a draw of participants - they can help painting ;) | 19:55 |
| *** yamamoto has quit IRC | 19:55 | |
| mtreinish | anteaya: well until recently, I had several people complain about it so I knuckled under and switched to something a bit easier on the eyes | 19:56 |
| fungi | the house is not subject to community covenants, so i could in theory paint it obnoxious fuchsia and the neighbors would have to put up rith it. but also this is the beach, so it wouldn't be that unusual | 19:56 |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Uncap pytz and add a note on ordering between Babel and pycadf https://review.openstack.org/182755 | 19:56 |
| openstackgerrit | Joshua Harlow proposed openstack-infra/project-config: Add futurist adoption project changes https://review.openstack.org/182808 | 19:57 |
| _david_ | SpamapS, Sure i'm, but what you mean by color? | 19:57 |
| *** dprince has quit IRC | 19:57 | |
| anteaya | mtreinish: no, your etherpad troll was something I relied on | 19:58 |
| anteaya | mtreinish: what colour are you going with now, so I can spot you in future? | 19:58 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard: Added explicit puppetlabs-apt version to vagrant bootstrap. https://review.openstack.org/182810 | 19:58 |
| *** Rockyg_ has quit IRC | 19:59 | |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config: Update team structure and add council https://review.openstack.org/182811 | 19:59 |
| mtreinish | anteaya: well, apparently I was scaring off contribs to the qa summit brainstorming etherpad because of the wall of red | 19:59 |
| *** craigbr has joined #openstack-infra | 19:59 | |
| fungi | mtreinish: i rather liked it | 19:59 |
| *** e0ne has joined #openstack-infra | 19:59 | |
| lifeless | SpamapS: how about supermarionation rainbow bikesheds? | 19:59 |
| fungi | then again, mine's probably equally obnoxious | 19:59 |
| mtreinish | anteaya: I switched to a more brownish red: https://etherpad.openstack.org/p/APtLUXk6Qk | 19:59 |
| *** nfedotov1 has quit IRC | 19:59 | |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone endpoint resource methods https://review.openstack.org/180328 | 19:59 |
| mtreinish | fungi: heh, I like your color choice | 20:00 |
| fungi | it does stand out | 20:00 |
| mtreinish | so at least we're consistent in our taste | 20:00 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone endpoint resource methods https://review.openstack.org/180328 | 20:01 |
| fungi | i just wish etherpad would feather the edges of the background so that it looked more like a grumbly black raincloud bubble whenever i type | 20:01 |
| *** Rockyg has joined #openstack-infra | 20:01 | |
| SpamapS | _david_: the whole back-story around bikeshedding is not agreeing on color and wasting time selecting one. | 20:01 |
| _david_ | SpamapS, OK, I see. | 20:02 |
| *** nfedotov has joined #openstack-infra | 20:03 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/requirements: Remove dev-requirements.txt https://review.openstack.org/182776 | 20:03 |
| *** Mmike has quit IRC | 20:05 | |
| anteaya | fungi: ha ha ha | 20:05 |
| anteaya | I want to see the bug report to etherpad | 20:06 |
| anteaya | mtreinish: you frightening person you | 20:06 |
| openstackgerrit | Doug Hellmann proposed openstack-infra/release-tools: Submit .gitreview to the right branch https://review.openstack.org/182818 | 20:07 |
| clarkb | I just approved the use shade change | 20:07 |
| * greghaynes drops the baloons | 20:08 | |
| fungi | (brace for impact!) | 20:08 |
| *** baoli has joined #openstack-infra | 20:08 | |
| pleia2 | yay | 20:08 |
| *** dannywilson has joined #openstack-infra | 20:09 | |
| lifeless | _david_: SpamapS: one more note there: its also about arguing over trivia being easy, vs adding real insight to a conversation | 20:09 |
| greghaynes | clarkb: oh, you got your node booting? | 20:09 |
| clarkb | greghaynes: ya itworked | 20:09 |
| greghaynes | clarkb: the qemu one | 20:09 |
| greghaynes | nice! | 20:09 |
| clarkb | yup | 20:09 |
| *** mattfarina has quit IRC | 20:09 | |
| lifeless | _david_: SpamapS: e.g. 'do we buy a truck or a motorcycle', lets make it red. | 20:09 |
| greghaynes | clarkb: vivit or trusty? | 20:09 |
| clarkb | greghaynes: trusty | 20:10 |
| greghaynes | clarkb: about how long did it take? and you were able to ssh in? | 20:10 |
| clarkb | so I am going to write up a devstack plugin for nodepool to run as a devstack service | 20:10 |
| *** bswartz has joined #openstack-infra | 20:10 | |
| clarkb | greghaynes: its about on par with the rax image. just under 400 seconds | 20:10 |
| *** eharney has joined #openstack-infra | 20:10 | |
| greghaynes | hrm, not bad | 20:10 |
| greghaynes | im gonna make some dib tests that do that, then | 20:10 |
| *** Mmike has joined #openstack-infra | 20:12 | |
| openstackgerrit | Merged openstack-infra/nodepool: Use shade to get the client objects https://review.openstack.org/168603 | 20:13 |
| clarkb | ok I think if I run puppet on nodepool.o.o now we get that version | 20:13 |
| clarkb | doing so | 20:13 |
| *** _david_ has quit IRC | 20:13 | |
| clarkb | (just speeding up the normal puppet agent run window) | 20:14 |
| *** mmedvede has quit IRC | 20:15 | |
| SpamapS | lifeless: I agree (thank god, otherwise we'd be bikeshedding about what bikeshedding is about) | 20:15 |
| clarkb | grr lxml was uninstalled for some reason and it tried to build after | 20:16 |
| clarkb | I don't even | 20:16 |
| *** esker has quit IRC | 20:16 | |
| mtreinish | jeblair: when you get a sec can you take a look at: https://review.openstack.org/181907 | 20:16 |
| *** carl_baldwin has joined #openstack-infra | 20:16 | |
| mtreinish | it's needed to plug a testing gap because of kilo | 20:17 |
| electrical | zaro: did you manage to get the details + fix for python-jenkins ? | 20:17 |
| clarkb | the global requirements lxml version is < that what we have installed | 20:17 |
| fungi | krotscheck: looks like the storyboard-infra-docs job is now known to zuul and the jenkins masters, so approve a change at will | 20:18 |
| zaro | electrical: i was going to ask pelix detail but he seems to have left the room | 20:19 |
| fungi | krotscheck: in theory some time soonish after it merges you should see content at http://docs.openstack.org/infra/storyboard/ | 20:19 |
| *** Mmike has quit IRC | 20:19 | |
| *** Mmike has joined #openstack-infra | 20:19 | |
| clarkb | it is because we do a pip install -U nodepool | 20:19 |
| krotscheck | fungi: Here we go: https://review.openstack.org/#/c/182042/ | 20:19 |
| *** mattfarina has joined #openstack-infra | 20:19 | |
| electrical | zaro: ah okay. i can create a gist for you with details if you want? | 20:19 |
| jeblair | mtreinish, anteaya: done, thanks | 20:19 |
| mtreinish | jeblair: cool, thanks | 20:19 |
| zaro | electrical: did you say someone has a fix ready to go? | 20:19 |
| clarkb | fungi: jeblair mordred pleia2 jhesketh SergeyLukjanov I think I am just going to install the dev headers it wants then rerun the install then update our puppet-nodepool manifest | 20:19 |
| clarkb | does that seem reasonable? | 20:19 |
| electrical | zaro: we do have a fix yeah, just not coded so you could merge it. but i can gist the fix yeah | 20:20 |
| fungi | clarkb: for lxml? yeah, seems fine | 20:20 |
| pleia2 | clarkb: yeah, I think that's fine | 20:20 |
| clarkb | yup so libxml2-dev and libxslt-dev | 20:20 |
| *** dustins has joined #openstack-infra | 20:20 | |
| zaro | electrical: why not just submit it to python-jenkins? | 20:20 |
| *** maishsk has quit IRC | 20:21 | |
| *** kmartin has quit IRC | 20:21 | |
| fungi | clarkb: for some reason i thought we had solved that already, but i suppose not | 20:21 |
| clarkb | fungi: we had, but not with pip install -U | 20:21 |
| *** amuller has left #openstack-infra | 20:21 | |
| fungi | grr | 20:21 |
| *** ildikov has joined #openstack-infra | 20:21 | |
| fungi | right | 20:21 |
| electrical | zaro: could do as well :-) | 20:21 |
| clarkb | fungi: with the -U pip is trying to upgrade to latest lxml via pypi | 20:21 |
| fungi | clarkb: yep, i get it. carry on ;_ | 20:21 |
| zaro | electrical: that would be preferable. thanks | 20:21 |
| electrical | zaro: okay. will do that most likely tomorrow. ( past 9 pm here ) | 20:22 |
| *** maishsk has joined #openstack-infra | 20:22 | |
| clarkb | jeblair: you ok with that plan? I want to say you had the preference for python-lxml via apt | 20:22 |
| openstackgerrit | Merged openstack-infra/storyboard: Revert "Email Send Cron Worker" https://review.openstack.org/182042 | 20:23 |
| clarkb | jeblair: it sounds like we should have lxml removed as a dep this cycle so I am happy to go ahead with this as the current plan | 20:24 |
| jeblair | clarkb: yeah, it's all a mess until we have packages. :( | 20:24 |
| fungi | krotscheck: 260a68f storyboard-infra-docs is now queued in the post pipeline | 20:24 |
| jeblair | i'm okay with your suggestion | 20:24 |
| krotscheck | fungi: yay! | 20:24 |
| *** claudiub has quit IRC | 20:24 | |
| *** pcrews has joined #openstack-infra | 20:25 | |
| clarkb | jeblair: thanks | 20:25 |
| *** abregman has quit IRC | 20:26 | |
| *** mmedvede has joined #openstack-infra | 20:27 | |
| clarkb | nodepool==0.1.2.dev1 is the version pip freeze reports now, getting the puppet-nodepool change written before I restart nodepool | 20:27 |
| *** david-lyle_ has joined #openstack-infra | 20:29 | |
| *** otter768 has joined #openstack-infra | 20:29 | |
| openstackgerrit | Merged openstack-infra/project-config: Add kilo compat jobs to tempest and devstack-gate https://review.openstack.org/181907 | 20:30 |
| *** mwagner_lap has quit IRC | 20:30 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/puppet-nodepool: Install lxml from source to accomodate pip -U https://review.openstack.org/182830 | 20:33 |
| clarkb | jeblair: fungi pleia2 ^ thats the puppet side of the change, I am ready to restart nodepool now, anything I should wait on? | 20:34 |
| *** julim has quit IRC | 20:34 | |
| *** otter768 has quit IRC | 20:34 | |
| *** marcusvrn1 has quit IRC | 20:35 | |
| *** marcusvrn has joined #openstack-infra | 20:35 | |
| pleia2 | I see the change, but pretty irc logs + making old ones pretty too is a new thing, right? | 20:36 |
| clarkb | pleia2: is that in response to me? not sure I have neough context to understand | 20:37 |
| *** doug-fish has left #openstack-infra | 20:38 | |
| jeblair | clarkb: nak | 20:38 |
| *** kmartin has joined #openstack-infra | 20:38 | |
| pleia2 | clarkb: no, had just noticed pretty irc logs made by eavesdrop | 20:38 |
| clarkb | oh, ya mordred added that late last week iirc | 20:38 |
| pleia2 | can now link to specific sections in the logs | 20:38 |
| *** freerunner has joined #openstack-infra | 20:38 | |
| *** annegentle has quit IRC | 20:39 | |
| clarkb | ok restarting nodepool now | 20:39 |
| *** annegentle has joined #openstack-infra | 20:39 | |
| lifeless | SpamapS: so meta, much shed | 20:40 |
| SpamapS | lifeless: wow | 20:40 |
| clarkb | its up, doing things including calculating demand and running tasks | 20:41 |
| *** sharikapongubala has joined #openstack-infra | 20:41 | |
| greghaynes | ooo | 20:41 |
| clarkb | next rax dib image uploads :) | 20:43 |
| greghaynes | clarkb: so looking good? | 20:43 |
| clarkb | greghaynes: ya so far I haven't seen anything bad in the logs | 20:43 |
| clarkb | will keep watching it though | 20:43 |
| *** erikmwilson has quit IRC | 20:43 | |
| *** bhunter71 has quit IRC | 20:43 | |
| openstackgerrit | Merged openstack/requirements: Allow Magnum to conform to global requirements process https://review.openstack.org/170830 | 20:44 |
| *** craigbr has quit IRC | 20:44 | |
| *** ajo has joined #openstack-infra | 20:47 | |
| openstackgerrit | greghaynes proposed openstack-infra/nodepool: Use shade for image uploads https://review.openstack.org/168633 | 20:48 |
| *** achanda has quit IRC | 20:48 | |
| greghaynes | clarkb: ^ so theres a caveat there about shade blocking on v1 uploads | 20:48 |
| greghaynes | not sure how much we care | 20:48 |
| greghaynes | oh, it should already be a blocking call | 20:49 |
| greghaynes | so not really a change | 20:49 |
| *** jamespd has joined #openstack-infra | 20:49 | |
| clarkb | ya existing should block | 20:49 |
| clarkb | I think each build runs in its own thread so blocking is fine | 20:49 |
| fungi | krotscheck: it's alive! http://docs.openstack.org/infra/storyboard/ | 20:50 |
| krotscheck | fungi: Brraaaaaaiiiinssss | 20:50 |
| *** maishsk has quit IRC | 20:50 | |
| fungi | anybody else want to +2/approve https://review.openstack.org/182790 ? it's been manually tested and works | 20:50 |
| jamespd | where does one report errors received on review.openstack.org? I'm getting a 'server error' when trying to set my Contact Information. | 20:51 |
| *** nfedotov has quit IRC | 20:51 | |
| fungi | jamespd: see if https://ask.openstack.org/question/56720 solves it | 20:51 |
| *** ildikov has quit IRC | 20:52 | |
| *** radez is now known as radez_g0n3 | 20:52 | |
| jamespd | fungi: ah, thanks! In retrospect, I should have been able to find that myself; soz. | 20:55 |
| clarkb | greghaynes: looking at https://review.openstack.org/#/c/168633/10/nodepool/fakeprovider.py where does FakeList.create() throw a runtime error? | 20:56 |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/system-config: Decoupe openstack_project::server from Gerrit https://review.openstack.org/182840 | 20:57 |
| clarkb | greghaynes: slightly worried the fake is trying to emulate shade but missing that piece | 20:57 |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Uncap pytz and block pycadf 0.6.2 https://review.openstack.org/182755 | 20:57 |
| mtreinish | can someone do me a favor and kick this: https://review.openstack.org/#/c/181747/ it doesn't want to go into the gate because of the depends-on (which merged a little while ago) | 20:57 |
| clarkb | mtreinish: looking | 20:58 |
| clarkb | approved | 20:58 |
| mtreinish | clarkb: normally when that happens I just +A it again, but I can't do that on project-config | 20:58 |
| mtreinish | clarkb: thanks | 20:59 |
| *** Sukhdev has joined #openstack-infra | 20:59 | |
| greghaynes | clarkb: oh, it should be changed now - shade should only throw a OpenstackCloudError | 20:59 |
| clarkb | greghaynes: right its catching a runtime error to through an openstackcloud error | 20:59 |
| clarkb | greghaynes: but I don't see anything that would trigger the runtime error | 21:00 |
| greghaynes | clarkb: yea, im not sure where thats from | 21:01 |
| greghaynes | clarkb: we should test that case though | 21:01 |
| *** Sukhdev_ has quit IRC | 21:01 | |
| *** bttully has joined #openstack-infra | 21:02 | |
| *** isviridov_away is now known as isviridov | 21:03 | |
| *** tiswanso has joined #openstack-infra | 21:03 | |
| *** miqui_ has joined #openstack-infra | 21:04 | |
| Clint | sdake: could you confirm that you want #kolla logged as well? | 21:04 |
| clarkb | greghaynes: looks like if we call create on the same object more than once then threading will raise RunTimeError | 21:04 |
| clarkb | greghaynes: so thats a case that should never happen if I am reading the tests correctly | 21:05 |
| *** isq_ has joined #openstack-infra | 21:05 | |
| *** kashyap` has joined #openstack-infra | 21:05 | |
| *** aysyanne has joined #openstack-infra | 21:05 | |
| *** dizquierdo has joined #openstack-infra | 21:06 | |
| *** shashankhegde_ has joined #openstack-infra | 21:06 | |
| *** markvan_ has joined #openstack-infra | 21:06 | |
| *** timfreun1 has joined #openstack-infra | 21:06 | |
| *** mriedem has quit IRC | 21:06 | |
| *** NobodyCa1 has joined #openstack-infra | 21:07 | |
| *** pfalleno1 has joined #openstack-infra | 21:07 | |
| *** GheRiver1 has joined #openstack-infra | 21:07 | |
| *** ildikov has joined #openstack-infra | 21:07 | |
| fungi | okay, rackspace has given us <24 hours (they don't say how much <) to reboot these https://etherpad.openstack.org/p/venom-reboots | 21:07 |
| *** arielb has quit IRC | 21:08 | |
| fungi | some we can just do immediately, some will take some coordination | 21:08 |
| fungi | i'm going to go through real fast and mark the ones we can do asap | 21:08 |
| pleia2 | sounds good | 21:08 |
| *** shardy has quit IRC | 21:08 | |
| *** mattymo_ has joined #openstack-infra | 21:09 | |
| clarkb | pleia2: one set that we can do gracefully with a small amount of admin work are the git* hosts. Any chance you are available to help with those and I can walk you through it? | 21:09 |
| *** ivar-lazzaro has quit IRC | 21:09 | |
| jeblair | fungi, pleia2: should we use the incident room for this? | 21:09 |
| *** wayneseguin has joined #openstack-infra | 21:09 | |
| *** dustins has quit IRC | 21:09 | |
| *** rakhmerov has quit IRC | 21:09 | |
| *** wayneeseguin has quit IRC | 21:09 | |
| *** mleroy has quit IRC | 21:09 | |
| *** kashyap has quit IRC | 21:09 | |
| *** tiswanso_ has quit IRC | 21:09 | |
| *** aysyd has quit IRC | 21:09 | |
| *** ldnunes has quit IRC | 21:09 | |
| *** ekarlso has quit IRC | 21:09 | |
| *** larsks has quit IRC | 21:09 | |
| *** cody-somerville has quit IRC | 21:09 | |
| *** wolsen has quit IRC | 21:09 | |
| *** mkj_wk_ has quit IRC | 21:09 | |
| *** sgordon has quit IRC | 21:09 | |
| *** NobodyCam has quit IRC | 21:09 | |
| *** freerunner has quit IRC | 21:09 | |
| *** shashankhegde has quit IRC | 21:09 | |
| *** markvan has quit IRC | 21:09 | |
| *** btully has quit IRC | 21:09 | |
| *** Kennan has quit IRC | 21:09 | |
| *** miqui has quit IRC | 21:09 | |
| *** pfallenop has quit IRC | 21:09 | |
| *** samueldmq has quit IRC | 21:09 | |
| *** greghaynes has quit IRC | 21:09 | |
| *** dmsimard has quit IRC | 21:09 | |
| *** GheRivero has quit IRC | 21:09 | |
| *** mattymo has quit IRC | 21:09 | |
| *** isq has quit IRC | 21:09 | |
| *** leifmadsen has quit IRC | 21:09 | |
| *** timfreund has quit IRC | 21:09 | |
| *** bttully is now known as btully | 21:09 | |
| *** freerunner has joined #openstack-infra | 21:09 | |
| *** greghayn1 has joined #openstack-infra | 21:09 | |
| *** shashankhegde_ is now known as shashankhegde | 21:09 | |
| *** mleroy has joined #openstack-infra | 21:09 | |
| *** ivar-lazzaro has joined #openstack-infra | 21:09 | |
| *** isq_ has quit IRC | 21:09 | |
| *** isq_ has joined #openstack-infra | 21:09 | |
| *** kashyap` has quit IRC | 21:09 | |
| *** kashyap` has joined #openstack-infra | 21:09 | |
| *** markvan_ has quit IRC | 21:10 | |
| *** markvan_ has joined #openstack-infra | 21:10 | |
| mtreinish | fungi: what about trove nodes? I would have assumed they're rebooting those too | 21:10 |
| *** wayneseguin is now known as wayneeseguin | 21:10 | |
| *** sgordon has joined #openstack-infra | 21:10 | |
| *** ldnunes has joined #openstack-infra | 21:10 | |
| *** samueldmq has joined #openstack-infra | 21:10 | |
| *** sgordon has quit IRC | 21:10 | |
| *** sgordon has joined #openstack-infra | 21:10 | |
| *** ekarlso has joined #openstack-infra | 21:10 | |
| *** wolsen has joined #openstack-infra | 21:10 | |
| *** ivar-lazzaro has quit IRC | 21:10 | |
| clarkb | mtreinish: thats a good question, in theory you can't break out of those nodes because you don't get shells on them | 21:10 |
| jeblair | (i don't have a strong opinion, just thought i'd throw it out there; we can also move there only if it gets too chatty here) | 21:10 |
| *** Kennan has joined #openstack-infra | 21:10 | |
| pleia2 | jeblair: I think so | 21:10 |
| *** rakhmerov has joined #openstack-infra | 21:10 | |
| *** leifmadsen has joined #openstack-infra | 21:10 | |
| clarkb | mtreinish: which probably means in theory that they don't consider those nodes compromised? | 21:10 |
| *** ivar-lazzaro has joined #openstack-infra | 21:11 | |
| fungi | jeblair: ooh! is it logged already? | 21:11 |
| *** dmsimard has joined #openstack-infra | 21:11 | |
| pleia2 | fungi: it is! | 21:11 |
| jeblair | fungi: it is, all ready to go! | 21:11 |
| *** larsks has joined #openstack-infra | 21:11 | |
| *** tiswanso has quit IRC | 21:11 | |
| *** NobodyCa1 is now known as NobodyCam | 21:11 | |
| *** cody-somerville has joined #openstack-infra | 21:11 | |
| fungi | mtreinish: no idea--they didn't say :/ | 21:11 |
| clarkb | mtreinish: or rather compromiseable | 21:11 |
| fungi | mtreinish: they might not be pvhvm? | 21:11 |
| *** achanda has joined #openstack-infra | 21:11 | |
| fungi | could be any number of raisins | 21:11 |
| mtreinish | clarkb: oh, yeah that's a good point. I guess if you're not allowed to shell into them it doesn't matter | 21:11 |
| clarkb | pleia2: the first thing we have to do with git* is take git02 or git01 out of the DNS round robin | 21:12 |
| openstackgerrit | Merged openstack-infra/project-config: Switch tempest python 3.4 unit test job to gating https://review.openstack.org/181747 | 21:13 |
| openstackgerrit | Clint 'SpamapS' Byrum proposed openstack-infra/shade: Handle novaclient exceptions during delete_server https://review.openstack.org/182845 | 21:13 |
| openstackgerrit | Clint 'SpamapS' Byrum proposed openstack-infra/shade: Handle novaclient exception in delete_server wait https://review.openstack.org/182846 | 21:13 |
| fungi | i agree on using the incident channel | 21:14 |
| *** mkj_wk has joined #openstack-infra | 21:14 | |
| jeblair | okay. To the incident root! | 21:14 |
| jeblair | room | 21:14 |
| *** craigbr has joined #openstack-infra | 21:14 | |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/requirements: Pin os-brick to the current release https://review.openstack.org/182756 | 21:14 |
| *** ldnunes has quit IRC | 21:15 | |
| *** ChanServ changes topic to "Active incident: venom related reboots | Related work in #openstack-infra-incident" | 21:16 | |
| *** che-arne has quit IRC | 21:17 | |
| *** dkranz has quit IRC | 21:19 | |
| *** gokrokve has quit IRC | 21:20 | |
| *** gokrokve has joined #openstack-infra | 21:21 | |
| *** kgiusti has quit IRC | 21:22 | |
| *** mriedem has joined #openstack-infra | 21:25 | |
| *** gokrokve has quit IRC | 21:25 | |
| *** miqui_ has quit IRC | 21:26 | |
| fungi | ttx: any chance you might be around? we're going to need to gracefully hard reboot design-summit-prep before rax helpfully reboots it for us | 21:29 |
| *** matrohon has joined #openstack-infra | 21:29 | |
| *** gokrokve has joined #openstack-infra | 21:29 | |
| *** greghayn1 is now known as greghaynes | 21:29 | |
| *** ibiris is now known as ibiris_away | 21:30 | |
| *** mrmartin has quit IRC | 21:31 | |
| mtreinish | fungi: I thought ttx said he was going to kill it today or tomorrow to lock down the sched | 21:31 |
| mtreinish | but I could just be making things up | 21:31 |
| fungi | okay, we might be in luck then | 21:31 |
| anteaya | mtreinish: any idea where he said that? | 21:31 |
| mtreinish | anteaya: it was either in the x-project meeting | 21:31 |
| mtreinish | or in an email to ptls | 21:32 |
| * mtreinish checks his inbox | 21:32 | |
| *** annegentle has quit IRC | 21:32 | |
| anteaya | I'll check the x-project meeting logs | 21:32 |
| openstackgerrit | Matt Riedemann proposed openstack/requirements: Uncap pytz https://review.openstack.org/182858 | 21:32 |
| mtreinish | anteaya: didn't see anything in my inbox so it was probably the meeting | 21:32 |
| *** angela-s has left #openstack-infra | 21:33 | |
| *** craigbr has quit IRC | 21:34 | |
| *** kmartin has quit IRC | 21:34 | |
| anteaya | mtreinish: I see this: http://eavesdrop.openstack.org/irclogs/%23openstack-meeting/%23openstack-meeting.2015-05-12.log.html#t2015-05-12T21:14:29 | 21:34 |
| anteaya | what is cheddar? | 21:35 |
| mtreinish | anteaya: cheddar is the tool to do the scheduling which is running on that host | 21:35 |
| mtreinish | anteaya: https://github.com/ttx/summitsched | 21:35 |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/system-config: Decouple openstack_project::server from Gerrit https://review.openstack.org/182840 | 21:36 |
| *** hdd has quit IRC | 21:36 | |
| asselin | anteaya, hi...sorry....what was the question from this morning? I lost my scrollback | 21:36 |
| anteaya | fungi: looks like we have a winner, ttx announced cheddar with be brought down at EOD wednesday in the x-project meeting | 21:37 |
| *** pc_m has quit IRC | 21:38 | |
| anteaya | asselin: it was about zuul and version numbers, I learned that pbr makes one up if you pull the tip of master | 21:38 |
| mtreinish | anteaya: ttx can totally see the future | 21:38 |
| anteaya | mtreinish: he is magic that way | 21:38 |
| fungi | anteaya: thanks. i'll just reboot it and let him sort it out tomorrow if he still needs it | 21:38 |
| anteaya | asselin: it bases it off of the last tag and increments by 0.0.1 | 21:38 |
| anteaya | fungi: sure | 21:38 |
| asselin | anteaya, I see | 21:39 |
| asselin | I also see http://zuul.openstack.org/ is not working today? | 21:39 |
| asselin | must be the same issue I ran into yesterday | 21:39 |
| anteaya | we we are in the midst of rebooting the world | 21:39 |
| asselin | or that :) | 21:39 |
| *** viglesias has quit IRC | 21:39 | |
| anteaya | so let's let them finish and then see what is up with zuul.o.o | 21:39 |
| anteaya | :) | 21:39 |
| fungi | or looks like the resolution is actually just to let ttx reboot it himself if he wants, otherwise allow rax to ungracefully power-cycle it (no idea if they try to halt it first) | 21:39 |
| *** ildikov has quit IRC | 21:40 | |
| anteaya | very good | 21:40 |
| anteaya | either way priority on contents seems low | 21:40 |
| fungi | btw i'm going to stop paying attention in here until we're wrapping up with the incident | 21:40 |
| asselin | fyi...here's the patch I had issues with. https://review.openstack.org/#/c/169956/ | 21:41 |
| anteaya | yup | 21:41 |
| asselin | for some reason I didn't get the new lib folder created...I guess it's just me.... | 21:41 |
| greghaynes | clarkb: so the RuntimeException is being raised by the dummy in fakeprovider | 21:42 |
| greghaynes | clarkb: had to spend a few to page it in | 21:42 |
| greghaynes | clarkb: also ignore me while youall incident | 21:42 |
| anteaya | asselin: you didn't get the new lib folder created, I don't see a new lib folder in that patch | 21:43 |
| anteaya | yet the commit message says it should be there | 21:43 |
| asselin | https://review.openstack.org/#/c/169956/1/etc/status/fetch-dependencies.sh | 21:43 |
| *** timfreun1 is now known as timfreund | 21:44 | |
| anteaya | thank you | 21:44 |
| openstackgerrit | greghaynes proposed openstack-infra/nodepool: Use shade for image uploads https://review.openstack.org/168633 | 21:44 |
| *** Sukhdev_ has joined #openstack-infra | 21:44 | |
| *** peristeri has quit IRC | 21:44 | |
| anteaya | so what happened you pulled in the merged version and didn't get the /lib folder created in your environment? | 21:44 |
| *** gokrokve has quit IRC | 21:45 | |
| *** viglesias has joined #openstack-infra | 21:45 | |
| *** gokrokve has joined #openstack-infra | 21:45 | |
| *** Krinkle|detached is now known as Krinkle | 21:46 | |
| SpamapS | hm | 21:46 |
| SpamapS | functional tests for shade failing | 21:46 |
| SpamapS | http://logs.openstack.org/45/182845/1/check/check-shade-dsvm-functional-neutron/402753e/console.html#_2015-05-13_21_34_47_258 | 21:46 |
| *** Sukhdev has quit IRC | 21:46 | |
| SpamapS | http://logs.openstack.org/45/182845/1/check/check-shade-dsvm-functional-neutron/402753e/console.html#_2015-05-13_21_34_47_258 | 21:47 |
| asselin | anteaya, correct....so zuul ui was still trying to get the files from the old location. | 21:47 |
| *** tnovacik has quit IRC | 21:47 | |
| asselin | anteaya, but I need to see if I have that script...perhaps there's an issue since I'm partially out-of-tree | 21:47 |
| anteaya | asselin: ah yes please check that first | 21:48 |
| anteaya | SpamapS: is there a difference between the two urls you posted? | 21:48 |
| anteaya | if yes, I can't see it | 21:48 |
| greghaynes | SpamapS: huh | 21:48 |
| greghaynes | SpamapS: novaclient regression? | 21:48 |
| *** amitgandhinz has quit IRC | 21:48 | |
| * greghaynes kicks off a tox -r | 21:49 | |
| SpamapS | anteaya: no, I meant to paste a snippet of the content | 21:49 |
| SpamapS | forgot to ctrl-c | 21:49 |
| anteaya | ah | 21:49 |
| SpamapS | shade.exc.OpenStackCloudException: Error getting compute endpoint: Error constructing auth plugin: password | 21:49 |
| anteaya | yes | 21:50 |
| *** markvan_ is now known as markvan | 21:50 | |
| anteaya | I know zip about shade | 21:50 |
| anteaya | so I have no insight apart from the obvious | 21:50 |
| *** gokrokve has quit IRC | 21:50 | |
| SpamapS | greghaynes: likely | 21:50 |
| SpamapS | greghaynes: released today | 21:50 |
| *** isviridov is now known as isviridov_away | 21:50 | |
| anteaya | and a new tox was released today as well | 21:51 |
| SpamapS | for extra great happy | 21:51 |
| anteaya | we didn't have enough happy apparently | 21:51 |
| *** puranamr has joined #openstack-infra | 21:51 | |
| SpamapS | it _totally_ helps that there are people on my roof hammering and tarring shingles, and a cleaning person vacuuming up the mess they made yesterday hammering. TOTALLY helping concentration. | 21:52 |
| anteaya | oh yes | 21:52 |
| anteaya | I can totally hear where you are coming from | 21:52 |
| openstackgerrit | James Polley proposed openstack-dev/pbr: WIP: Teach pbr to read extras from setup.cfg https://review.openstack.org/181795 | 21:53 |
| mtreinish | anteaya: oh, based on that changelog the new tox it will likely break some people | 21:53 |
| *** gasgasgas has joined #openstack-infra | 21:53 | |
| gasgasgas | Win 20.000-30.000 Euro Per year (1700-2500 Euro per month) With 1 H per day for 365 days. At start you will earn little money but with time your starting to learn much more at start you can win 0.5 euro per day first week and after a month you will earn 10-20 euro per day Its Verry simple just Make an account on my link and i will train you and i will be your guide as much you | 21:53 |
| gasgasgas | need Sign Up here And good Luck -----> http://www.marketglory.com/strategygame/lolopoco | 21:53 |
| anteaya | mtreinish: yes it has | 21:53 |
| *** ildikov has joined #openstack-infra | 21:53 | |
| anteaya | lifeless: ^^ | 21:53 |
| mtreinish | anteaya: they added env variable isolation | 21:53 |
| anteaya | lifeless: I don't have ops | 21:53 |
| greghaynes | seems legit | 21:53 |
| lifeless | anteaya: ? I don't think I do either. | 21:54 |
| anteaya | you do I saw in teh manifest, give it a shot | 21:54 |
| lifeless | k, sec while I look up the incantation | 21:54 |
| anteaya | thanks | 21:54 |
| lifeless | anteaya: doesn't look like it | 21:56 |
| lifeless | 09:55 <lifeless> akick #openstack-infra add gasgasgas!*@* | 21:56 |
| lifeless | 09:55 -ChanServ(ChanServ@services.)- You are not authorized to perform this operation. | 21:56 |
| anteaya | :( | 21:56 |
| lifeless | can't op myself either. | 21:56 |
| openstackgerrit | fumihiko kakuma proposed openstack-infra/puppet-jenkins: Add retry parameters for 'jenkins-jobs update' https://review.openstack.org/182868 | 21:56 |
| anteaya | k | 21:56 |
| JayF | lifeless: you're in the access list | 21:56 |
| JayF | lifeless: maybe not identified to nickserv? | 21:56 |
| lifeless | I'm identified, I log in with sasl :) | 21:56 |
| JayF | Heh, awesome :) | 21:56 |
| JayF | strange it won't work for you then | 21:57 |
| lifeless | but just in case | 21:57 |
| Clint | lifeless: i think you got unidentified | 21:57 |
| *** ChanServ sets mode: +b gasgasgas!*@* | 21:57 | |
| *** gasgasgas was kicked by ChanServ (User is banned from this channel) | 21:57 | |
| anteaya | thank you | 21:57 |
| pleia2 | np | 21:57 |
| *** ChanServ sets mode: +o lifeless | 21:57 | |
| lifeless | deop lifeless | 21:58 |
| lifeless | Clint: yeah, thats where I'd just got to :) | 21:58 |
| *** lifeless sets mode: -o lifeless | 21:58 | |
| * Clint nods. | 21:58 | |
| *** ajo has quit IRC | 21:58 | |
| greghaynes | SpamapS: hrm, tox -r isnt failing for me | 21:58 |
| SpamapS | greghaynes: functional | 21:58 |
| SpamapS | greghaynes: it's not run by default | 21:58 |
| greghaynes | oh | 21:58 |
| *** mriedem is now known as mriedem_away | 21:58 | |
| greghaynes | well, there ya go | 21:59 |
| * SpamapS is re-stacking now | 21:59 | |
| SpamapS | my devstack test instance has been up for 327 days | 21:59 |
| SpamapS | that seems.. wrong. ;) | 21:59 |
| greghaynes | SpamapS: it is annoying we swallow the internal traceback | 21:59 |
| mordred | SpamapS: btw - if you're stacking ... | 21:59 |
| mordred | oh - nm - you're on it already | 21:59 |
| SpamapS | mordred: functional fails in shade? | 21:59 |
| mordred | yah | 22:00 |
| SpamapS | that is why I'm stacking. :) | 22:00 |
| SpamapS | mordred: novaclient released today. | 22:00 |
| SpamapS | seems.. likely :) | 22:00 |
| mordred | SpamapS: oh. yay | 22:00 |
| * SpamapS prepares to bisect | 22:00 | |
| mordred | well - we ALSO released a new os-cient-config | 22:00 |
| SpamapS | triple win | 22:00 |
| SpamapS | (tox release today too) | 22:00 |
| mordred | clarkb: did your nodepool test work? | 22:00 |
| fungi | reed: we're going to need to reboot groups.o.o and ask.o.o shortly. should be brief. any reason we should wait? does anyone need notifying? | 22:01 |
| greghaynes | mordred: he said he got a qemu node booting and able to ssh in | 22:01 |
| reed | fungi, give me a sec | 22:01 |
| mordred | greghaynes: great. so there isn't a general problem with current shade and current os-client-config - but there may be something screwy in our tests | 22:01 |
| mordred | johnthetubaguy: I know you're probably busy, what with the general carnage - but if you happen to be around | 22:02 |
| mordred | johnthetubaguy: I'm rebooting all these nodes, and one of them is in 'powering-off' state | 22:02 |
| mordred | johnthetubaguy: and has been for some time ... do I need to file a rax support ticket? | 22:02 |
| fungi | reed: sure, no problem | 22:02 |
| reed | fungi, go | 22:02 |
| *** aysyanne has quit IRC | 22:03 | |
| fungi | reed: thanks! | 22:03 |
| *** carl_baldwin has quit IRC | 22:03 | |
| fungi | reed: ask is on its way back up now | 22:05 |
| clarkb | mordred: it did work and nodepool is running with it | 22:05 |
| mordred | clarkb: oh! it's live? neat | 22:05 |
| clarkb | mordred: yup | 22:05 |
| mordred | clarkb: good to know we got something accomplished before today's uberbug | 22:06 |
| *** soren_ has quit IRC | 22:06 | |
| mordred | BASED ON FLOPPY DISKS | 22:06 |
| *** gordc has quit IRC | 22:06 | |
| fungi | reed: groups should be coming up now too | 22:06 |
| mordred | fungi, reed: have we put in a questions about openstackid? | 22:06 |
| fungi | mordred: not yet | 22:07 |
| fungi | reed: groups is looking good to me, but ask seems ill | 22:07 |
| fungi | troubleshooting now | 22:07 |
| reed | oh no | 22:07 |
| reed | mordred, it depends on the question :) | 22:07 |
| *** rlandy has quit IRC | 22:08 | |
| mordred | reed: well, we need to reboot it - so the question is "anything special we need to do or special care that needs to be taken?" | 22:08 |
| reed | mordred, oh... now I get it :) | 22:09 |
| reed | I think it can be rebooted quickly without problems | 22:09 |
| reed | what happened? | 22:09 |
| fungi | reed: ask seems to have recovered on its own | 22:09 |
| fungi | just takes it a bit coming up i suppose | 22:09 |
| reed | i see internal server error | 22:09 |
| reed | will try again | 22:09 |
| reed | same with another browser | 22:10 |
| mordred | reed: same thing with ask and groups - we're going through and rebooting all of our servers because of CVE | 22:10 |
| reed | ah | 22:10 |
| reed | reboot all the things | 22:11 |
| *** gokrokve has joined #openstack-infra | 22:11 | |
| *** achanda has quit IRC | 22:13 | |
| *** hdd has joined #openstack-infra | 22:14 | |
| *** eharney has quit IRC | 22:14 | |
| *** salv-orlando has joined #openstack-infra | 22:15 | |
| *** salv-orlando has quit IRC | 22:15 | |
| *** matrohon has quit IRC | 22:17 | |
| *** bhunter71 has joined #openstack-infra | 22:18 | |
| mordred | reed: done. openstackid has also been rebooted | 22:18 |
| SpamapS | ok... functional tests passing with old novaclient.. trying new | 22:18 |
| reed | mordred, openstackid looks good | 22:19 |
| *** jamielennox|away is now known as jamielennox | 22:19 | |
| SpamapS | pass with python-novaclient 2.25 | 22:20 |
| *** bknudson has quit IRC | 22:20 | |
| greghaynes | hrm | 22:20 |
| *** julim has joined #openstack-infra | 22:20 | |
| SpamapS | and os-client-config 1.0.0 | 22:20 |
| SpamapS | _might_ be server side | 22:20 |
| SpamapS | I didn't git pull everything | 22:21 |
| *** hashar has quit IRC | 22:22 | |
| *** ivar-lazzaro has quit IRC | 22:23 | |
| *** sabeen2 has quit IRC | 22:23 | |
| mordred | SpamapS greghaynes: totally not important - but we can also remove that unset OS_CACERT now | 22:23 |
| mordred | as shade will properly ignore that it's not there | 22:23 |
| SpamapS | mordred: yay | 22:23 |
| SpamapS | mordred: I have a patch up to stop setting it to a non-existant file in devstack too | 22:24 |
| mordred | SpamapS: neat! | 22:24 |
| *** ivar-lazzaro has joined #openstack-infra | 22:24 | |
| SpamapS | since that's actually a legitimate complaint IMO | 22:24 |
| mordred | SpamapS: btw - if you get bored - I keep thining that we should get devstack to emit a clouds.yaml | 22:24 |
| SpamapS | "Hey you told me to load CA certs here and there's no CA certs here." | 22:24 |
| *** ivar-lazzaro has quit IRC | 22:24 | |
| greghaynes | ooo | 22:25 |
| clarkb | SpamapS: there is a specific comment in devstack that refutes that complaint | 22:25 |
| *** ivar-lazzaro has joined #openstack-infra | 22:25 | |
| clarkb | SpamapS: its says "this is fine if tls is not enabled" | 22:25 |
| *** nelsnelson has quit IRC | 22:25 | |
| mordred | yup. otoh, writing a cert to the file when you didn't create the cert is just messy | 22:25 |
| *** ivar-lazzaro has quit IRC | 22:25 | |
| SpamapS | clarkb: yes, that was the assumption that was wrong which I have corrected. :) | 22:26 |
| *** baoli has quit IRC | 22:26 | |
| mordred | and if part of devstack's job is to document things, then the OS_CACERT setting could certainly confuse someone | 22:26 |
| mordred | even if it doesn't actively break something | 22:26 |
| *** ivar-lazzaro has joined #openstack-infra | 22:26 | |
| SpamapS | It's pretty normal to have pre-flight checks on configuration values. | 22:27 |
| *** mattfarina has quit IRC | 22:27 | |
| *** e0ne has quit IRC | 22:28 | |
| clarkb | I dunno, the only confusion I had was that shade would care about that value when not using https, but thats been fixed so I am happy | 22:28 |
| SpamapS | ok, so functional tests pass with everything new in the venv. I'm re-pulling everything and re-stacking. | 22:28 |
| SpamapS | clarkb: there's an argument that if a security setting is present, it should be handled in fail-closed mode, as in "You said CACERT, you didn't mean it, but I'm going to fail in case you screwed that up" | 22:29 |
| SpamapS | it's pedantry, and I don't necessarily want to be pedantic, but the argument exists and isn't just in my head. | 22:30 |
| clarkb | SpamapS: except thats not really what happened | 22:30 |
| clarkb | SpamapS: what happened was I said use HTTP not HTTPS but you failed because you checked something completely unrelated to what is going on | 22:30 |
| *** erikmwilson_ is now known as erikmwilson | 22:30 | |
| clarkb | SpamapS: if however https was in use then absolutely fail | 22:30 |
| *** dannywilson has quit IRC | 22:30 | |
| *** otter768 has joined #openstack-infra | 22:30 | |
| SpamapS | clarkb: The two things are conflicting. The fact that the URL has a higher precedence is not exactly "well thought out". | 22:31 |
| clarkb | uh | 22:31 |
| SpamapS | It's not important. | 22:31 |
| clarkb | SpamapS: the CACERT has no value unless using the protocol | 22:31 |
| clarkb | SpamapS: so its perfectly well thought out | 22:31 |
| SpamapS | yes, but if it is present, there's a clue that you may have meant for the protocol to be https. | 22:31 |
| *** freerunner has quit IRC | 22:32 | |
| SpamapS | It's not important, it's just one way to handle sensitive configuration. | 22:32 |
| clarkb | SpamapS: in that case your failure mode was wrong anyways | 22:32 |
| reed | fungi, ask.openstack.org seems to be responding correctly now | 22:32 |
| *** dannywilson has joined #openstack-infra | 22:32 | |
| clarkb | SpamapS: the failure should be "https is not in use but CACERT is set" | 22:32 |
| clarkb | SpamapS: that is completely different than what shade was doing | 22:32 |
| SpamapS | clarkb: yes, I agree 100%. If this pedantry is intentional, then the error messages have to be clear. It was accidental. | 22:33 |
| tchaypo | If the user specifies http but *also* specifies https, I don’t think it’s reasonable to assume that the user intended http | 22:33 |
| *** Sukhdev has joined #openstack-infra | 22:33 | |
| tchaypo | it seems more likely to me that the user was confused | 22:33 |
| *** patrickeast has quit IRC | 22:33 | |
| *** patrickeast_ has joined #openstack-infra | 22:33 | |
| *** Rockyg has quit IRC | 22:33 | |
| fungi | reed: yeah, it was fine about 3-4 minutes after it rebooted, i was just impatient | 22:33 |
| tchaypo | I don’t think that erroring is “you checked something completely unrelated”, it sounds to me like E_PEBKAC | 22:33 |
| fungi | reed: worth noting it returns an internal server error for a few minutes after a reboot | 22:34 |
| *** ekarlso has quit IRC | 22:34 | |
| *** ekarlso has joined #openstack-infra | 22:34 | |
| SpamapS | tchaypo: I'm ok with there being precedence too. OS_AUTH_URL having http or not is perfectly fine superseding the implied setting that is "OS_CACERT isn't set therefore I use system certs" | 22:34 |
| *** dimtruck is now known as zz_dimtruck | 22:34 | |
| reed | fungi, yes, probably because apache comes up before the rest of the machinery behind | 22:34 |
| greghaynes | SpamapS: what do you think about making our shade OpenStackCloudError also print out the internal exception? | 22:35 |
| *** otter768 has quit IRC | 22:35 | |
| greghaynes | SpamapS: Im worried that in production were going to end up having a lot of pain around debugging errors | 22:35 |
| greghaynes | er, also print out the internal stacktrace | 22:36 |
| tchaypo | Yeah, I think having a precendence is fine too. Either way could make sense to me depending on the context | 22:36 |
| tchaypo | I can’t see either of them as “you failed because you checked something completely unrelated" | 22:36 |
| tchaypo | “you didn’t do what I wanted because your priorities are different from what I expected” perhaps | 22:36 |
| SpamapS | greghaynes: debug logging does that? | 22:36 |
| *** mattfarina has joined #openstack-infra | 22:36 | |
| clarkb | tchaypo: the old error message is what you are arguing against | 22:36 |
| fungi | reed: that does seem to be the case, yes | 22:36 |
| clarkb | tchaypo: it basicaly say "CACERT doesn't exist go away" | 22:37 |
| *** Sukhdev_ has quit IRC | 22:37 | |
| greghaynes | SpamapS: hrm. This is another "we shouldnt have to run prod with debug logging" hole. | 22:37 |
| jogo | do we have a shared doc on running unit tests with tox and whatnot? | 22:37 |
| SpamapS | greghaynes: might be a case-by-case thing. | 22:38 |
| mordred | greghaynes: all OpenStackCloudExceptions should be raising with exc_info on if they're re-raising | 22:38 |
| SpamapS | greghaynes: in general I'd like for us to set a useful message and raise that. | 22:38 |
| mordred | if they're not, it's a bug | 22:38 |
| mordred | no? | 22:38 |
| SpamapS | true | 22:39 |
| greghaynes | mordred: all the ones I see str(e) when they make the raised exception | 22:39 |
| *** ashleighfarnham has quit IRC | 22:39 | |
| -openstackstatus- NOTICE: Gerrit and Zuul are going offline for reboots to fix a security vulnerability. | 22:39 | |
| mordred | oh - log.debug has the exc_info | 22:39 |
| *** ChanServ changes topic to "Gerrit and Zuul are going offline for reboots to fix a security vulnerability." | 22:39 | |
| mordred | sigh | 22:39 |
| * mordred shuts up and goes to reboot nodes | 22:39 | |
| *** ashleighfarnham has joined #openstack-infra | 22:39 | |
| *** dims_ has joined #openstack-infra | 22:40 | |
| *** pabelanger has joined #openstack-infra | 22:40 | |
| *** mattfarina has quit IRC | 22:40 | |
| jogo | because every repo's developer doc seems to have something different | 22:40 |
| *** achanda has joined #openstack-infra | 22:40 | |
| SpamapS | greghaynes: we could do this http://stackoverflow.com/questions/3847503/wrapping-exceptions-in-python | 22:40 |
| greghaynes | SpamapS: Yes, that is exactly what im thinking | 22:41 |
| anteaya | jogo I created this a long time ago: https://review.openstack.org/#/c/33456/8/TESTING.rst | 22:41 |
| jogo | and i just hit a 503 on gerrit :/ | 22:41 |
| anteaya | jogo: and gerrit is down now so give it a minute | 22:41 |
| *** dims has quit IRC | 22:41 | |
| anteaya | timing ftw | 22:41 |
| jogo | anteaya: :) | 22:41 |
| greghaynes | SpamapS: I think in a perfect world we would have two cases - known exceptions where we should have a good enough message from the client to debug and unkown exceptions where we log both stacktraces above debug | 22:42 |
| anteaya | the openstack rickroll | 22:42 |
| greghaynes | SpamapS: but were a ways from that | 22:42 |
| anteaya | jogo: here it is: http://git.openstack.org/cgit/openstack/oslo-incubator/tree/TESTING.rst | 22:42 |
| SpamapS | greghaynes: logging stacktraces in prod is a last resort.. especially for a library. | 22:42 |
| jogo | anteaya: so very nice | 22:43 |
| anteaya | jogo: thanks | 22:43 |
| *** Sukhdev has quit IRC | 22:43 | |
| anteaya | I have a strong feeling noone uses it | 22:43 |
| jogo | anteaya: but was hoping to get something under docs.openstack.org/developer/ or something | 22:43 |
| anteaya | however it does exist | 22:43 |
| *** dims has joined #openstack-infra | 22:43 | |
| anteaya | jogo: people might find it there | 22:43 |
| jogo | and nova has this http://docs.openstack.org/developer/nova/devref/unit_tests.html | 22:43 |
| *** teran has joined #openstack-infra | 22:43 | |
| anteaya | they can't find it where it is | 22:43 |
| SpamapS | greghaynes: I'd rather see us say that if we don't know what the exception is, we'll wrap it. | 22:44 |
| SpamapS | greghaynes: but logging should remain a debug log. | 22:44 |
| anteaya | yup, same info | 22:44 |
| greghaynes | SpamapS: Yes, so idea - can we add a property to OpenStackCloudError that is the internal stack, and then in the repr well print both | 22:44 |
| greghaynes | then its up to the using app whether they care | 22:44 |
| *** ashleighfarnham has quit IRC | 22:44 | |
| anteaya | jogo: discoverability is the factor, common areas, like a public park or square | 22:44 |
| SpamapS | greghaynes: something like that yeah | 22:45 |
| *** claudiub has joined #openstack-infra | 22:45 | |
| jogo | anteaya: I wonder if we can get that moved to http://docs.openstack.org/infra/manual/developers.html | 22:46 |
| jogo | assuming that is the correct home | 22:46 |
| anteaya | jogo: offer a patch and we both find out | 22:46 |
| jogo | anteaya: do you know what repo that is in? | 22:46 |
| anteaya | oslo-incubator is the source, infra-manual is the destination | 22:47 |
| *** dims_ has quit IRC | 22:47 | |
| *** openstackgerrit has quit IRC | 22:47 | |
| *** mattfarina has joined #openstack-infra | 22:48 | |
| *** openstackgerrit has joined #openstack-infra | 22:49 | |
| *** claudiub has quit IRC | 22:49 | |
| *** claudiub has joined #openstack-infra | 22:50 | |
| *** spzala has quit IRC | 22:50 | |
| pabelanger | Are the zuul / gerrit security issues in the host OS or the applications themselves? | 22:52 |
| Clint | kernel bug | 22:53 |
| pabelanger | roger | 22:53 |
| anteaya | pabelanger: http://seclists.org/oss-sec/2015/q2/425 | 22:54 |
| *** tonytan4ever has quit IRC | 22:54 | |
| pabelanger | Ah | 22:54 |
| pabelanger | anteaya, I seen ttx tweet that. I lol'd are the marketing too | 22:54 |
| anteaya | so this was initiated by our host and we are responding to their decision | 22:54 |
| anteaya | well that is a legitamate question | 22:55 |
| *** david-lyle_ has quit IRC | 22:55 | |
| anteaya | as it seems folks will exploit any emotion to further their business goals | 22:55 |
| anteaya | but we are responding to our host's decision here | 22:55 |
| anteaya | we didn't choose to do this | 22:56 |
| *** asselin_ has joined #openstack-infra | 22:57 | |
| jhesketh | Morning | 22:57 |
| anteaya | morning jhesketh | 22:57 |
| anteaya | fun in progress | 22:57 |
| *** ChanServ changes topic to "Active incident: venom related reboots | Related work in #openstack-infra-incident" | 22:57 | |
| -openstackstatus- NOTICE: Gerrit and Zuul are back online. | 22:57 | |
| *** dannywilson has quit IRC | 22:58 | |
| *** patrickeast_ has quit IRC | 22:58 | |
| *** patrickeast has joined #openstack-infra | 23:01 | |
| *** asselin_ has quit IRC | 23:01 | |
| *** dannywilson has joined #openstack-infra | 23:01 | |
| *** Rockyg has joined #openstack-infra | 23:07 | |
| *** achanda has quit IRC | 23:09 | |
| openstackgerrit | Joshua Harlow proposed openstack-infra/project-config: Add futurist adoption project changes https://review.openstack.org/182808 | 23:09 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add Keystone service resource methods https://review.openstack.org/177621 | 23:09 |
| openstackgerrit | Davide Guerri proposed openstack-infra/shade: Add keystone endpoint resource methods https://review.openstack.org/180328 | 23:10 |
| *** ianw has joined #openstack-infra | 23:13 | |
| *** panda has quit IRC | 23:13 | |
| *** gokrokve has quit IRC | 23:13 | |
| *** gokrokve has joined #openstack-infra | 23:14 | |
| *** panda has joined #openstack-infra | 23:14 | |
| *** gokrokve has quit IRC | 23:15 | |
| *** signed8bit is now known as signed8bit_ZZZzz | 23:16 | |
| *** aduarte has quit IRC | 23:16 | |
| *** ChanServ changes topic to "Discussion of OpenStack Developer and Community Infrastructure | docs http://docs.openstack.org/infra/manual/ http://ci.openstack.org/ | bugs https://storyboard.openstack.org/ | source https://git.openstack.org/cgit/openstack-infra/" | 23:16 | |
| fungi | okay, my half dozen held nodes are now deleted | 23:17 |
| clarkb | fwiw I am double checking on the logging log situation and it looks like those nodes are special because they are the pvhvm with only 20GB / | 23:17 |
| fungi | there are two remaining held nodes, which i will assume are clarkb's | 23:17 |
| clarkb | so I think thats mostly cleaned up and will write a note to myself to write logrotate rules for these logs so we stop doing that | 23:17 |
| clarkb | fungi: ya I have two, I will delte the one I don't want | 23:17 |
| clarkb | done | 23:18 |
| clarkb | pleia2: looks like DNS is back to normal for git.o.o too :) | 23:18 |
| fungi | i had held representatives of each corresponding devstack and bare node in each provider for bindep pre/post package list comparison testing | 23:18 |
| jeblair | centos6 nodes are ready | 23:18 |
| fungi | wrapped that up earlier this week | 23:18 |
| pleia2 | clarkb: cool, I'll bump the ttl back up to a reasonable level in a bit | 23:19 |
| greghaynes | clarkb: updated https://review.openstack.org/#/c/168633/11 btw | 23:19 |
| clarkb | pleia2: I think 5 minutes is what we want for the round robin | 23:20 |
| clarkb | pleia2: but maybe jeblair/fungi can say otherwise | 23:20 |
| *** vipuls is now known as vipul | 23:20 | |
| pleia2 | wfm | 23:20 |
| openstackgerrit | Joe Gordon proposed openstack-infra/infra-manual: Add ChangeLog and AUTHORS to .gitignore https://review.openstack.org/182878 | 23:21 |
| openstackgerrit | Joe Gordon proposed openstack-infra/infra-manual: Add Running Unit Tests section to developers doc https://review.openstack.org/182879 | 23:21 |
| clarkb | I dunno I have had DNS round robins with health checks that ran on 30 second tlls | 23:21 |
| jogo | anteaya: ^ there | 23:21 |
| openstackgerrit | Steve Baker proposed openstack/requirements: Add sphinxcontrib-blockdiag https://review.openstack.org/182880 | 23:21 |
| clarkb | I am sure that breaks all the DNS rules but it DNS is bad at failover | 23:21 |
| fungi | clarkb: pleia2: yeah, 5 minutes is fine. that's what they were already at before we started | 23:21 |
| fungi | for a round-robin pool i think that's a reasonable choice anyway since you want to be able to take members in and out of it relatively quickly | 23:22 |
| clarkb | ya | 23:22 |
| pleia2 | yeah | 23:22 |
| fungi | if you think 30 seconds is short, f5 used to recommend for their 3dns product that you set cache times to 0 or 1 second | 23:23 |
| clarkb | ES is slowly but surely recovering | 23:23 |
| clarkb | I will keep an eye on the health of that this evening | 23:23 |
| openstackgerrit | Debo~ Dutta proposed openstack-infra/project-config: WIP: Cognitive - Machine Learning as a Service https://review.openstack.org/182886 | 23:24 |
| *** ivar-lazzaro has quit IRC | 23:24 | |
| fungi | however, for the longest time i remember at least a couple of well-known broadband providers that would on their customer-facing recursive resolvers reqrite any ttl less than a day | 23:24 |
| fungi | er, rewrite | 23:24 |
| *** ivar-lazzaro has joined #openstack-infra | 23:24 | |
| pleia2 | packing up for debian jessie release party in a bit (we have jessie party hats!) but I'm around for a bit longer | 23:24 |
| fungi | cowgirl hats? | 23:24 |
| clarkb | greghaynes: and I never got to debugging dib git cache things | 23:25 |
| clarkb | greghaynes: but we are using shade now so its not all bad | 23:25 |
| pleia2 | fungi: nah, the cone ones that have a picture of jessie on them :) | 23:25 |
| fungi | oh, cool enough anyway! | 23:25 |
| greghaynes | clarkb: yes, I really want to dig into debugging that | 23:25 |
| *** ivar-lazzaro has quit IRC | 23:25 | |
| greghaynes | clarkb: ETOOMANYTHINGS | 23:25 |
| *** ivar-lazzaro has joined #openstack-infra | 23:25 | |
| fungi | i had several things i was planning to do this afternoon, but now that the sun's down i think i'm going to wind down too | 23:26 |
| jeblair | fungi: follow the sun | 23:26 |
| fungi | on saturday i will, for sure | 23:26 |
| *** achanda has joined #openstack-infra | 23:26 | |
| *** Daviey has quit IRC | 23:28 | |
| fungi | hrm, puppetboard says all the zuul servers (including mergers) are failing to puppet. ask.o.o too | 23:28 |
| *** david-lyle_ has joined #openstack-infra | 23:28 | |
| fungi | i wonder if that puppet-zuul change from earlier today is to blame for the former | 23:28 |
| *** mattfarina has quit IRC | 23:28 | |
| fungi | yeah, looks like maybe it's trying to make files absent when their parent directory doesn't exist? | 23:29 |
| fungi | that's a rather strange failure mode, but then again, it's puppet | 23:30 |
| *** signed8bit_ZZZzz is now known as signed8bit | 23:31 | |
| *** hdd has quit IRC | 23:31 | |
| anteaya | jogo: commented | 23:31 |
| jeblair | jogo, anteaya: i'm not certain that's the best place for that | 23:32 |
| *** mattfarina has joined #openstack-infra | 23:32 | |
| clarkb | greghaynes: you should be able to use our git caching element on a image build, boot that, then run zuul-cloner against it to reproduce | 23:33 |
| clarkb | greghaynes: thats basically what my plan was | 23:33 |
| clarkb | greghaynes: my hunch is that GitPython doing a clone locally does something similar to an rsync rather than doing a clone and setting up the remote refs properly | 23:33 |
| clarkb | greghaynes: but it does so in an incomplete way so that the fetch_remote thing is wrong | 23:33 |
| *** JayF has quit IRC | 23:33 | |
| anteaya | jeblair: great, since we can't figure out a discoverable place for it | 23:34 |
| jeblair | infra-manual is "the user manual for the project infrastructure", and this is getting into some project-specific stuff. it's largely the same because we made it the same for openstack projects, but, for instance, it's focused heavily on tox which is only useful for python projects. | 23:34 |
| anteaya | it has been in olso-incubator for 2 years now and I'm sure noone finds it | 23:34 |
| *** mattfarina has quit IRC | 23:34 | |
| *** banix has quit IRC | 23:34 | |
| jeblair | anteaya: i don't think the text in oslo-incubator was intendet to be a projcet wide reference | 23:34 |
| *** Daviey has joined #openstack-infra | 23:35 | |
| anteaya | well I certainly wanted people to find that file | 23:35 |
| fungi | jhesketh: https://review.openstack.org/172007 has a problem. see http://puppetboard.openstack.org/report/zuul.openstack.org/b0ff213b478d2c9e57f5e7b077e757e5ca9c7145 | 23:35 |
| anteaya | and followed the advice I was given | 23:35 |
| *** JayF has joined #openstack-infra | 23:35 | |
| jeblair | anteaya: i think the best place we have for something like that is the wiki | 23:36 |
| *** Krinkle is now known as Krinkle|detached | 23:37 | |
| jeblair | anteaya: or within projects themselves | 23:37 |
| fungi | jhesketh: aha, i see the problem. patch on the way | 23:37 |
| anteaya | jogo: ^^ help yourself | 23:37 |
| jhesketh | fungi: ah thanks for letting me know | 23:37 |
| jhesketh | fungi: oh you beat me to it | 23:37 |
| *** mattfarina has joined #openstack-infra | 23:37 | |
| jeblair | anteaya, jogo: the use of "tox" specifically is a red flag for me | 23:39 |
| *** mattfarina has quit IRC | 23:39 | |
| *** wenlock has quit IRC | 23:39 | |
| *** Krinkle|detached is now known as Krinkle | 23:40 | |
| jeblair | this is kind of a grey area -- we do talk about bugs and blueprints, but they interact with gerrit via jeepyb, and that's intended to be universal. tox is too, to a certain extent, but is still only relevant for python projects, etc, and it's hard to generalize about what tests might be in a project | 23:40 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/puppet-zuul: Make sure /var/lib/zuul/www/lib exists https://review.openstack.org/182890 | 23:40 |
| fungi | jhesketh: ^ | 23:40 |
| anteaya | I've let it grow mould for 2 years now and am content to continue | 23:41 |
| mordred | jeblair: I could certainly imagine the existence of a thing like infra-manual except that is full of things like that devref | 23:41 |
| jhesketh | fungi: thanks, lgtm | 23:41 |
| mordred | because we know how putting reference info into the wiki winds up going | 23:42 |
| mordred | random people decide to add "help" | 23:42 |
| jeblair | mordred: well, i'm not sure what's wrong with TESTING files in projects | 23:43 |
| mordred | but given that we're talking about interacting with the Project Testing Interface, it does seem like OpenStack-wide documentation suitable for something | 23:43 |
| mordred | jeblair: because divergence and copying | 23:43 |
| jeblair | mordred: divergence may be necessary | 23:43 |
| mordred | jeblair: it may be - but there is also a ton of openstack-specific things that a dev needs to know that has solid non-changing answers | 23:44 |
| jeblair | the text as proposed to infra-manual is a series of processes one might undertake to learn how one might run tests that may or may not be any given project. if the project happens to be in python | 23:44 |
| jeblair | wheras the TESTING file in zuul indicates exactly how one should run tests for zuul | 23:44 |
| *** rfolco has quit IRC | 23:45 | |
| mordred | we've spent an awful lot of effort in aligning the projects so that I think for non-infra per-project testing files should be considered a bug | 23:45 |
| jeblair | mordred: i agree this may be a useful thing, but i'm uncertain who the audience is | 23:45 |
| mordred | jeblair: people who are working in opensatck for the first time | 23:45 |
| asselin | any cores want to +A this spec? https://review.openstack.org/#/c/135170/ | 23:45 |
| jeblair | mordred: that's not true -- we certainly permit extra tests | 23:45 |
| *** ociuhandu has quit IRC | 23:45 | |
| mordred | jeblair: we do - but we try not to permit extra test running mechanisms | 23:46 |
| asselin | CI Dashboard Spec | 23:46 |
| *** teran_ has joined #openstack-infra | 23:46 | |
| jeblair | mordred: i think pycadf's testing file could easily support saying "you should run these tox environments under these circumstances" | 23:46 |
| jeblair | since it has like 50 of them | 23:46 |
| mordred | I would think that a per-project TESTING file should indicate specialness or divergence from the norm | 23:46 |
| *** yamamoto has joined #openstack-infra | 23:47 | |
| mordred | but we shouldn't have 200 copies of "to run python unittests, run tox -epy27" | 23:47 |
| jesusaurus | jeblair: clarkb: I ran some tox tests on some various zuul commits and put the results at https://etherpad.openstack.org/p/76057-timings but its hard to draw any conclusions from that since not many tests are duplicated across the different slow lists | 23:47 |
| *** dizquierdo has quit IRC | 23:48 | |
| *** teran has quit IRC | 23:48 | |
| jesusaurus | though we definitely have a trend of tests getting slower, both over time and with 76057 being applied | 23:48 |
| jeblair | mordred: i don't agree that is bad -- i think it's fine for a project to have things like TESTING, README, and CONTRIBUTING which help the project stand on its own. i also think TESTING could pretty easily link to a wiki page or some other central documentation | 23:48 |
| *** bradm_ is now known as bradm | 23:48 | |
| jeblair | mordred: so i propose the latter as a compromise positon :) | 23:48 |
| mordred | sure. I can see that. I could also see a slightly more wordy canonical description of how the PTI works and what you should expect to do with it as the content of the thing linked to from the TESTING files? | 23:49 |
| mordred | jeblair: :) | 23:49 |
| *** ajmiller has quit IRC | 23:49 | |
| jeblair | mordred: so here are the reasons i don't want it in infra-manual: it's language/repo specific, it's not required to interact with the project infrastructure, and it's not directly tied to the project infrastructure | 23:50 |
| mordred | totally grok those an agree | 23:50 |
| mordred | and | 23:50 |
| *** hichihara has joined #openstack-infra | 23:50 | |
| mordred | I'm arguing that there should probably be somewhere that somethign liek this should live | 23:50 |
| mordred | I don't know where | 23:50 |
| *** sharikapongubala has quit IRC | 23:50 | |
| mordred | but I don't think it's a wiki - because of all the reasons | 23:51 |
| jeblair | mordred: yeah, let's see if we can find a place | 23:51 |
| openstackgerrit | Clark Boylan proposed openstack-infra/puppet-logstash: Logrotate the indexer.log logfile. https://review.openstack.org/182894 | 23:51 |
| clarkb | any chance I can get quick review on ^ | 23:51 |
| mordred | NO | 23:51 |
| mordred | jk | 23:51 |
| clarkb | that fixes aprt of the logging system logs filling disk problem | 23:51 |
| *** bknudson has joined #openstack-infra | 23:51 | |
| boris-42 | mordred: hi | 23:51 |
| mordred | hi boris-42 | 23:51 |
| clarkb | the harder one is ES because its rotating on its own just not cleaning up, might be easiest to use a find to delete older logs there | 23:51 |
| boris-42 | mordred: I see some NOT-REGISTER jobs | 23:51 |
| jeblair | mordred: so ttx, i, and others are writing a guide for new projects; to the extent that we want to tell new projects "you should use the PTI and test running should generally work this way" we can add some of it there | 23:51 |
| boris-42 | mordred: is it ok? | 23:51 |
| clarkb | can logrotate easily not rotate and just delete logs? | 23:52 |
| mordred | fungi: ^^ you were looking at NOT-REGISTERED weren't you? | 23:52 |
| jeblair | boris-42: should be good now, try recheck | 23:52 |
| boris-42 | jeblair: so one more left I think recheck will help | 23:53 |
| jesusaurus | clarkb: ive had past experiences where logrotate wasn't deleting the logs because the service itself was rotating them, but that may have been poor logrotate config on my part | 23:53 |
| jesusaurus | clarkb: but i suspect that rotate+delete are tightly coupled in logrotate | 23:54 |
| fungi | mordred: jeblair: anteaya: maybe test design and development/testing standards are best published in a similar way and place as the api working group's recommendations? | 23:54 |
| fungi | from a cross-project recommendations/standards perspective anyway | 23:54 |
| jeblair | mordred: i don't have a suggestion for a developer-oriented place to put this. i'm open to discussing how to incorporate it into the manual, but part of the reason i wrote the manual is that people kept adding "optional extra things that might help" to the page which was supposed to be "here's what you need to know to be able to submit a change" | 23:55 |
| jogo | jesusaurus: I am not sure infra manual is the best place either | 23:55 |
| *** yamamoto has quit IRC | 23:55 | |
| anteaya | I stopped caring some time ago, I wrote the file there is is anyone is welcome to use it as they see fit | 23:55 |
| clarkb | jogo: ya I am thinking just cron a find with age rule and delete | 23:55 |
| boris-42 | jeblair: okay this one doesn't work gate-rally-install-bare-precise | 23:55 |
| clarkb | er jesusaurus ^ | 23:55 |
| jesusaurus | clarkb: ++ | 23:55 |
| pleia2 | clarkb: I think if you use "rotate 0" in the config and use find inline it won't rotate | 23:55 |
| pleia2 | clarkb: but at that point, you might as well just use find in cron :) | 23:55 |
| jogo | jeblair: what I do think is duplicating the docs per project isn't great and neither is a wiki | 23:55 |
| jogo | jeblair: I do see how adding python specific stuff to the general developer manual isn't great, but there is already some stuff in there subtly (not saying that makes it better) | 23:56 |
| jogo | jeblair: what about a python specific page on some of this stuff? | 23:57 |
| jeblair | jogo: yeah, i know, i'm struggling with where to draw the line :) | 23:57 |
| jeblair | jogo: you might be on to something | 23:57 |
| *** mtanino has quit IRC | 23:57 | |
| jeblair | jogo: so maybe if, where you put that text, it was instead, "hey you might want to run some tests, see these links for relevent infos", and they could be links to howtos in the manual | 23:57 |
| jogo | jeblair: I don't really care where it lives as long as we don't have 20 copies of it that are each slightly different. and I just don't like wikis | 23:58 |
| jeblair | jogo: that would get it out of the critical path of "how do i submit a patch because this isn't github" | 23:58 |
| jeblair | which is my biggest worry with that doc | 23:58 |
| fungi | if we do a howtos section of the manual, the current sandbox tutorial page probably belongs there | 23:58 |
| SpamapS | hrm | 23:59 |
| jeblair | yeah, i feel similarly about that | 23:59 |
| fungi | though arguably, the new project creation steps are very howto as well | 23:59 |
| jeblair | it's useful content but does not need to be in the critical path | 23:59 |
| *** ashleighfarnham has joined #openstack-infra | 23:59 | |
| SpamapS | greghaynes: so even with latest devstack + latest tip of everything, not getting the failure we see in shade func tests | 23:59 |
| jeblair | fungi: true, though they have their own "guide" section | 23:59 |
| jeblair | jogo: i will update my review with that | 23:59 |
| * SpamapS tries pip installing the exact versions | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!