*** sdake_ has joined #openstack-infra | 00:00 | |
*** craigbr has quit IRC | 00:00 | |
*** MarkAtwood has joined #openstack-infra | 00:00 | |
fungi | perhaps the bug is clearer feedback that "this change depends on an abandoned or missing change N" rather than just "this change failed to merge" (or worse, silence) | 00:00 |
---|---|---|
lifeless | If Depends-On is analogous to child-of-git-commit, then I'd expect analogous behaviour to abandon on the parent - e.g. it gets blocked. Visibly so. | 00:00 |
jeblair | lifeless: it makes things quite complex if change X is on more than one branch or project. if you change your mind about whether X is appropriate for a given project-branch, then the only way to recover from that is to abandon all instances of change X, and re-upload them as new ones. | 00:00 |
lifeless | jeblair: you mean when someone uses the same Change-Id in a patch to many projects? | 00:01 |
jeblair | lifeless: i don't think it's ever okay to have a workflow where the process to correct an error is to lose work history. | 00:01 |
jeblair | lifeless: yes and/or the same change-id in a patch to many branches. | 00:01 |
clarkb | jeblair: we could also fix this by using truly unique keys | 00:02 |
* dims_ reads scroll back to see if anyone has seen "AttributeError: 'module' object has no attribute 'loader'" | 00:02 | |
clarkb | jeblair: then force each depends on to be explicit | 00:02 |
lifeless | dims_: that sounds like fallout from the new testtools release. | 00:02 |
dims_ | lifeless: yep | 00:02 |
lifeless | dims_: can you get me more details ? | 00:02 |
jeblair | clarkb: that means we have to develop new keys, and also generally, i think the behavior with shared change-ids is desirable. | 00:02 |
dims_ | lifeless - https://jenkins04.openstack.org/job/gate-oslo.config-python26/141/console | 00:02 |
lifeless | dims_: specifally I need to know the unittest2 version in use, and the traceback2 version | 00:02 |
dims_ | lifeless: looking | 00:03 |
lifeless | jeblair: so I can see that when there are 3 copies of a key, your concern makes sense. When there is only one, mine makes sense I think. | 00:03 |
clarkb | jeblair: or require a project,branch,changeid tuple as those are unique | 00:03 |
*** bhunter71 has quit IRC | 00:03 | |
lifeless | jeblair: perhaps only ignore abandoned when at least one other instance of the same key has landed. | 00:03 |
jeblair | clarkb: i suppose we could _extend_ the syntax to also support a more unique id as a recovery method for the case in point. i'm not sure it's worth it though. | 00:03 |
lifeless | jeblair: or even only look for a single copy of the id to have landed? | 00:03 |
*** ghostpl_ has joined #openstack-infra | 00:04 | |
jeblair | lifeless: i think "ignore abandoned except require at least one" could be a reasonable compromise behavior | 00:04 |
lifeless | jeblair: correct me if I'm wrong, but isn't another way to recover to remove that particular Depends-On as well? If everything necessary has in fact landed. | 00:04 |
jeblair | lifeless: i think the "all instances of X" have landed is particularly useful for the backport case, so i'd like to preserve it. | 00:04 |
dims_ | lifeless: traceback2-1.4.0-py2.py3-none-any.whl | 00:05 |
lifeless | jeblair: I can see the potential there. | 00:05 |
jeblair | lifeless: yes, that's true, you could manually delete the depends-on in that case. i suppose the issue comes up when you still want to depend on X_2 though you have abandoned X_1. | 00:05 |
jeblair | lifeless: or don't want to have to remember to clean that up later | 00:06 |
*** oomichi has quit IRC | 00:06 | |
dims_ | lifeless: unittest2==1.0.0 i think | 00:06 |
lifeless | dims_: I can't tell :/ | 00:07 |
lifeless | dims_: in fact, I think its not installed at all | 00:07 |
lifeless | dims_: it looks like extras failed to install | 00:07 |
jeblair | clarkb: I'd like to avoid "Depends-On: openstack-infra/nodepool~master~I2166355c44ea353869f5f4fb63e9c75cd567ab2f" if we can avoid it, but i could probably be convinced to support that _in addition_ to "Depends-On: I2166355c44ea353869f5f4fb63e9c75cd567ab2f" | 00:07 |
lifeless | ok | 00:08 |
clarkb | jeblair: ya I think the current behavior is a good default but supporting the more specific info might help corner cases | 00:08 |
lifeless | so extras pulls in testtools.TestCommand | 00:08 |
lifeless | which can't be imported at all until unittest2 is installed | 00:08 |
dims_ | lifeless: all python26 jobs | 00:08 |
clarkb | fungi: for the new ask site can you double check the uids and gids? curious if it breaks there too | 00:08 |
clarkb | fungi: when you uild the node (I think you said ou wold do that?) | 00:08 |
jeblair | i think my order of preference for this is a) "ignore abandoned except require at least one" and b) openstack-infra/nodepool~master~I2166355c44ea353869f5f4fb63e9c75cd567ab2f | 00:08 |
lifeless | its going to blow up everything in short order | 00:08 |
lifeless | the try_import would catch an ImportError | 00:09 |
lifeless | but not an AttributeError | 00:09 |
lifeless | working up a patch now - be five minutes | 00:09 |
dims_ | lifeless: thanks | 00:09 |
*** sdake__ has joined #openstack-infra | 00:10 | |
*** mtanino has quit IRC | 00:12 | |
*** mmedvede has quit IRC | 00:12 | |
lifeless | https://bugs.launchpad.net/testtools/+bug/1430076 dims_ | 00:14 |
openstack | Launchpad bug 1430076 in testtools "greenfield install error" [Undecided,New] | 00:14 |
*** sdake_ has quit IRC | 00:14 | |
*** andreykurilin_ has quit IRC | 00:14 | |
*** mmedvede has joined #openstack-infra | 00:16 | |
*** mmedvede has quit IRC | 00:17 | |
openstackgerrit | Steve Leon proposed openstack-infra/project-config: Adding new project: cue-dashboard to stackforge https://review.openstack.org/162804 | 00:18 |
fungi | clarkb: will do. i deleted my test instance but will launch a proper one now that the patch has landed | 00:19 |
fungi | er, once it merges i guess | 00:20 |
*** ujuc has joined #openstack-infra | 00:23 | |
lifeless | ok just waiting on travis | 00:26 |
lifeless | then we'll have this fixed in a couple minutes | 00:26 |
dims_ | lifeless: thanks | 00:28 |
lifeless | https://github.com/testing-cabal/testtools/pull/130 | 00:28 |
*** bhuvan has joined #openstack-infra | 00:28 | |
*** woodster_ has joined #openstack-infra | 00:29 | |
*** achanda has quit IRC | 00:30 | |
lifeless | ok 1.6.1 is up on PyPI | 00:30 |
*** yamamoto has quit IRC | 00:30 | |
lifeless | dims_: ^ | 00:31 |
*** zz_dimtruck is now known as dimtruck | 00:31 | |
dims_ | cool. will recheck and see | 00:31 |
*** hdd_ has joined #openstack-infra | 00:31 | |
*** tnovacik has quit IRC | 00:33 | |
openstackgerrit | Merged openstack-infra/system-config: Initial commit of Ask website https://review.openstack.org/140043 | 00:34 |
*** baoli has quit IRC | 00:34 | |
*** MarkAtwood has quit IRC | 00:36 | |
*** hichihara has joined #openstack-infra | 00:36 | |
*** hichihara has quit IRC | 00:36 | |
*** dannywil_ has quit IRC | 00:37 | |
*** camunoz has quit IRC | 00:40 | |
*** baoli has joined #openstack-infra | 00:40 | |
*** camunoz has joined #openstack-infra | 00:41 | |
StevenK | pleia2: Just a note, but I switched to trying to shoehorn Zanata into CentOS instead, based on the (wrong) thought, that since it was already packaged, it would be easy. I'm trying to break circular build depends at the moment. | 00:41 |
*** SumitNaiksatam has joined #openstack-infra | 00:42 | |
*** ChuckC has quit IRC | 00:42 | |
*** gyee has quit IRC | 00:44 | |
*** bdpayne has quit IRC | 00:45 | |
*** MarkAtwood has joined #openstack-infra | 00:45 | |
*** MarkAtwood has quit IRC | 00:46 | |
*** ddieterly has joined #openstack-infra | 00:46 | |
*** ChuckC has joined #openstack-infra | 00:49 | |
*** r-daneel has quit IRC | 00:49 | |
*** ddieterl_ has joined #openstack-infra | 00:51 | |
*** ChuckC has quit IRC | 00:53 | |
*** ChuckC has joined #openstack-infra | 00:53 | |
*** ddieterly has quit IRC | 00:54 | |
*** tiswanso has joined #openstack-infra | 00:55 | |
*** yamamoto has joined #openstack-infra | 00:55 | |
*** tiswanso has quit IRC | 00:56 | |
*** tiswanso has joined #openstack-infra | 00:57 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/requirements: Bump version of oslo.utils to 1.4.0 https://review.openstack.org/162644 | 00:58 |
fungi | clarkb: 104.239.140.123 though other than creating an admin group at gid 3000 i don't think it really exercised the potential race either way | 00:59 |
lifeless | ok subunit-1.1 released | 00:59 |
clarkb | fungi: thanks | 00:59 |
clarkb | also next time I should script this node cleanup | 01:00 |
clarkb | was a mistake to do this manually but its good lazy end of day work | 01:00 |
openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Move openstack_project::server to node def https://review.openstack.org/162819 | 01:03 |
mordred | ok. what have I missed | 01:09 |
mordred | anything good? | 01:09 |
clarkb | mordred: I think everyone is mostly trying to catch up on last week :) | 01:09 |
anteaya | I saw turkeys flying today | 01:09 |
anteaya | I rarely see turkeys flying | 01:10 |
anteaya | I saw four today | 01:10 |
anteaya | and saw their whole flight | 01:10 |
clarkb | mordred: I have shade functional tests running against the change that adds them | 01:10 |
clarkb | er queued in zuul at least | 01:10 |
*** melwitt has quit IRC | 01:11 | |
*** ujuc has quit IRC | 01:13 | |
*** rfolco has joined #openstack-infra | 01:13 | |
anteaya | are you at the ops meetup? | 01:13 |
anteaya | are the ops meeting up? | 01:13 |
*** jyuso1 has joined #openstack-infra | 01:13 | |
clarkb | I am not but I think mordred is | 01:14 |
mtreinish | clarkb: awesome thanks a bunch. I'll probably push the release when I get back from the ops midcycle | 01:14 |
*** baoli has quit IRC | 01:14 | |
anteaya | clarkb: yes sorry that was for mordred | 01:15 |
anteaya | mtreinish: how is it going thus far? | 01:15 |
anteaya | how many are there? | 01:15 |
*** sputnik13 has quit IRC | 01:15 | |
clarkb | I think I got through the bulk of outstanding puppet-* reviews today | 01:15 |
mtreinish | a ton of people I think they said ~180 | 01:15 |
clarkb | if I missed any do let me know | 01:15 |
anteaya | mtreinish: any idea if sdague has presented the nova-net to neutron migration status yet? | 01:16 |
mordred | clarkb: woot! I like it when you have functional tests!!! | 01:17 |
anteaya | the meeting happens in less than 8 hours would be great to share any responses from ops if I know what they are | 01:17 |
mordred | anteaya: and yeah - there are TONS of people | 01:17 |
mtreinish | anteaya: no I don't believe that's come up yet | 01:17 |
clarkb | mordred: well the change to have the tests in shade isn't merged yet :) we shall see how it does | 01:17 |
mordred | anteaya: pretty much everyone here runs neutron already | 01:17 |
anteaya | tons, 180 | 01:17 |
anteaya | okay | 01:17 |
anteaya | mordred: oh wonderful what a relief | 01:17 |
anteaya | mordred: oh okay thanks | 01:18 |
mordred | anteaya: I mean, we dida show of hands thing that went into the etherpad | 01:18 |
mordred | but I'd say at least 75% of the room was already running neutron, if not more | 01:18 |
anteaya | do you have a link to the etherpad? | 01:18 |
mordred | yeah - one sec | 01:18 |
anteaya | that is good information, thanks | 01:18 |
*** asselin_ has joined #openstack-infra | 01:19 | |
mordred | gah. mtreinish do you remember the link to the top level list? | 01:19 |
dstufft | I was gonna go to that ops thing, till my MIL decided she couldn't pick up my daughter from the bus >:[ | 01:19 |
mordred | dstufft: :( | 01:19 |
mordred | dstufft: well, I'm waving right now from the Westin | 01:19 |
*** asselin_ has quit IRC | 01:19 | |
anteaya | dstufft: :( | 01:19 |
anteaya | dstufft: you so rarely go to things | 01:19 |
lifeless | dstufft: MIL ? | 01:19 |
mtreinish | mordred: no I haven't had links to any of the things all day, but I'll dig it up one sec | 01:19 |
dstufft | lifeless: mother in law | 01:20 |
anteaya | sorry you couldn't go to this | 01:20 |
lifeless | dstufft: ah | 01:20 |
anteaya | mordred mtreinish thanks | 01:20 |
*** otter768 has joined #openstack-infra | 01:20 | |
dstufft | anteaya: the truth is I'm actually just a simulation | 01:20 |
mtreinish | anteaya, mordred: https://etherpad.openstack.org/p/PHL-ops-tags | 01:20 |
anteaya | dstufft: see this is where I don't believe you | 01:20 |
mordred | yes | 01:20 |
*** vilobhmm1 has quit IRC | 01:20 | |
anteaya | dstufft: mordred is a simulation, he disappears regularly | 01:20 |
anteaya | dstufft: you show up when stuff goes bad, see I can't buy your story | 01:21 |
anteaya | mtreinish: thanks | 01:21 |
mordred | anteaya: look for: 'inner ring' poll - expectation is that these components work together at a particular version | 01:21 |
mordred | anteaya: in that etherpad | 01:21 |
dstufft | anteaya: heh :) | 01:21 |
anteaya | mordred: ah thanks | 01:23 |
anteaya | dstufft: :) | 01:23 |
anteaya | I see a fan in attendance: How did some of these exit incubation? Based on what criteria? | 01:24 |
dstufft | (Maybe I'll wake up early enough to make it worth it to uber down in the morning and just come back with enough time to get Alyssa myself) | 01:25 |
mordred | anteaya: well, I mean, that's one of the reasons we reworked that system | 01:25 |
mordred | anteaya: beacuse they exited incubation because they conformed to process | 01:25 |
dstufft | I'm running on little sleep though so idk :/ | 01:25 |
clarkb | dstufft: you just need to clone yourself | 01:26 |
anteaya | dstufft: sounds like self-reliance is the ticket | 01:26 |
openstackgerrit | Ramy Asselin proposed openstack-infra/nodepool: Fix Nodepool DIB Uploads to Glance w/o region_name https://review.openstack.org/162313 | 01:26 |
clarkb | dstufft: then send dstufft' off on errands | 01:26 |
anteaya | mordred: well according to operators they don't even use or want what we currently _have_ in the openstack namespace | 01:26 |
anteaya | sorry much of what we have | 01:26 |
mordred | anteaya: another feedback from them is that they do not care if it's in the openstack namespace or not | 01:27 |
mordred | they care whether it works and whether it solves one of their needs | 01:27 |
anteaya | how do I get stuff tagged inner ring in review so I can review that stuff | 01:27 |
*** zhiwei has joined #openstack-infra | 01:28 | |
mordred | anteaya: oh - "inner ring" isn't a thing | 01:28 |
anteaya | my nervous system is currently doing deer in the headlights | 01:29 |
mordred | I think someone just called that quick straw-poll we took of the people in attendence that | 01:29 |
anteaya | I'd really like to work on things people want to use | 01:29 |
mordred | yes | 01:29 |
mordred | so - there are some takeaways from that room about getting some of the fokls in that room to work on providing us some more direct feedback about that | 01:29 |
anteaya | the etherpad is direct enough for me | 01:29 |
anteaya | unless folks who said zero usage were not being honest | 01:30 |
mordred | I asked "who is running XXX" | 01:30 |
mordred | anteaya: and zero people raised their hands | 01:30 |
mordred | in a room of 180 | 01:30 |
*** dimtruck is now known as zz_dimtruck | 01:30 | |
anteaya | okay thanks | 01:30 |
anteaya | that helps give me some context | 01:31 |
clarkb | mordred: are you saying there won't be a set of things that we test together as the "core" and tag appropriately? | 01:31 |
anteaya | to the percentages is your eyeball of the hands you saw | 01:31 |
mordred | I'm not saying that at all | 01:31 |
mordred | this was a discussion with the operators about what things we could do that would be useful to them | 01:31 |
clarkb | mordred: sure but you just said there is no inner ring thing | 01:31 |
mordred | there were many feedbacks | 01:31 |
anteaya | I'm all for figuring out how to be useful | 01:31 |
anteaya | since I've been very very busy | 01:31 |
mordred | clarkb: there is currently zero thing actually called "inner ring" | 01:31 |
anteaya | but I question how useful I'm being | 01:31 |
clarkb | mordred: I thought that was one of the things that we intended to tag simply because it helps the dev side | 01:31 |
mordred | clarkb: I think we're trying to figure out what would actually be helpful to people to tag | 01:32 |
mordred | thus the discussion | 01:32 |
mordred | like, you know - just ask the operators what they want :) | 01:32 |
lifeless | ok, -> lunch. shout if the subunit release borked anything. | 01:32 |
clarkb | mordred: but the tags aren't just for them aiui | 01:32 |
mordred | lifeless: what if I just want to shout? | 01:32 |
anteaya | I'd like some tag that says this is the group of things many operators are using | 01:32 |
clarkb | mordred: their input is good but its not the only input | 01:32 |
mordred | clarkb: nope. but we had them ina room | 01:32 |
mtreinish | lifeless: you know I will :) | 01:32 |
mordred | so we figured we ask them | 01:32 |
mordred | to see what we could learn | 01:32 |
mordred | clarkb: it was actually really helpful in some unexpected ways | 01:33 |
clarkb | thats good | 01:33 |
anteaya | mordred: do expand | 01:33 |
mordred | clarkb: for instance, it turns out global-requirements is WAY more useful to them than we might have expected before | 01:33 |
anteaya | what unexpected ways? | 01:33 |
anteaya | how so? | 01:33 |
mtreinish | mordred: well except I don't think anyone was actually thinking of how you do staged upgrades in the global reqs model, but whatever | 01:34 |
mordred | many of them build packages based on the single contigyous set of depends that the various openstack services need - and they dont' want version-per-service - because they want to have an easier time tracknig security bugs through their deployments | 01:34 |
clarkb | mtreinish: you don't :) | 01:34 |
mordred | I don't think they mind a moment in time when there are two versions | 01:34 |
mordred | they mind a long term divergent versions strategy | 01:35 |
waynr | greghaynes: ++ at decoupling the scheduler from changesets, I had started to get the feeling that something would have to change in the scheduler but haven't had time to fully grok what all is going on there | 01:35 |
clarkb | mordred: well with global reqs you can't really have two version globally | 01:35 |
mordred | for manageability | 01:35 |
mordred | clarkb: right | 01:35 |
clarkb | mordred: because pip and pkg_resources | 01:35 |
mordred | yu | 01:35 |
mordred | yup | 01:35 |
mordred | they're fans of that | 01:35 |
sdague | oh... that's the piece that I forgot to counter in there | 01:35 |
*** Daisy_ has joined #openstack-infra | 01:35 | |
clarkb | mordred: ya I don't think its a bad idea | 01:35 |
sdague | which is how do they expect to roll upgrades in such and environment | 01:35 |
mordred | most of them also seemed to be running releases | 01:35 |
clarkb | mordred: but it is a limitation that they may not expect | 01:35 |
mordred | and not master | 01:35 |
mordred | so they're only doing rolling upgrades every six months or so | 01:35 |
mordred | in which case, building a new set of depends, and then doing a rolling upgrade is kinda a known pattern | 01:36 |
clarkb | mordred: well you can't do a rolling upgrade | 01:36 |
mtreinish | mordred: sure but that's actually where the problem is because there often isn't overlap between the releases | 01:36 |
waynr | greghaynes: you mind if I take a first pass at the spec or do you think we should wait until changesets you mentioned are closer to ready? | 01:36 |
mordred | so - short story ... | 01:36 |
clarkb | mordred: thats pretty much not supported with this model | 01:36 |
clarkb | mordred: its an upgrade everything at least on one node all at once | 01:36 |
mordred | they all REALLY REALLY loudly said that the global reqs tied across the services was very important | 01:36 |
waynr | i can begin looking at those changesets with my reviewer's eye | 01:36 |
sdague | clarkb: yes, I can attest to mordred's statement about that | 01:37 |
greghaynes | waynr: go for it | 01:37 |
clarkb | sdague: mordred I am not disagreeing :) | 01:37 |
mordred | :) | 01:37 |
mordred | clarkb: I know | 01:37 |
mordred | I'm just trying to highlight how vocal it was | 01:37 |
clarkb | sdague: mordred but its important to know that these are not rolling upgrades and that in general they are not possible | 01:37 |
sdague | however, I do think the upgrade question wasn't asked | 01:37 |
clarkb | if you go with this model | 01:37 |
sdague | right | 01:37 |
mtreinish | I was really surprised by that too | 01:37 |
mtreinish | (that they were so vocal about it) | 01:37 |
lifeless | follow up q for them... how many of you deploy any java or ruby or javascript apps ? | 01:37 |
*** zhiwei has left #openstack-infra | 01:38 | |
sdague | I'm going to just ask jesse about it tomorrow, he was one of the vocal ones | 01:38 |
lifeless | cause none of them have this :) | 01:38 |
sdague | lifeless: none of which? | 01:38 |
dstufft | mordred: do they literally mean the global-requirements.txt mechanism, or do they mean the ability to install everything into one environment? | 01:38 |
lifeless | the native packaging systems for those three platforms allow multiple versions of deps to be present at once | 01:38 |
lifeless | sdague: ^ | 01:39 |
clarkb | dstufft: as I parse it everything in one env with common deps | 01:39 |
mordred | dstufft: they mean it is very important to them that all of openstack work with a single consistent set of requirements | 01:39 |
clarkb | dstufft: since they build packages anyways the mechanism is less important | 01:39 |
sdague | lifeless: right, gems are great that way | 01:39 |
mordred | they very actively do not want keystone and nova to need different versions of something | 01:39 |
sdague | it's a problem that's actually pretty python specific | 01:39 |
lifeless | sdague: so e.g. you can often end up with 5 or 6 different versions of X present at once. In your JVM or interpreter. | 01:39 |
mordred | lifeless: none of them are installing language packaging systems it did not seem | 01:39 |
dstufft | pkg_resources allows you to have multiple versions installed too fwiw | 01:39 |
mordred | lifeless: most of them are building debs or rpms and installing from that | 01:40 |
*** Ryan_Lane has quit IRC | 01:40 | |
dstufft | only one per process though | 01:40 |
clarkb | dstufft: how do you import them? | 01:40 |
mordred | right. nobody is doing that | 01:40 |
mordred | they're all installing from debs/rpms | 01:40 |
clarkb | dstufft: eg how do I say from foo import version1.6 | 01:40 |
mordred | which don't allow that | 01:40 |
mordred | and they're all pretty happy about that part | 01:40 |
dstufft | clarkb: your __main__ has a __requires__, or ther'es a pkg_resources api for it | 01:40 |
mordred | they just don't want us to make them have to deal with not being able to do that sanely | 01:40 |
dstufft | or you use a pkg_resources entry point and it's all automagically handled for you | 01:40 |
mordred | dstufft: well, except openstack overrides those :) | 01:41 |
dstufft | so does pip for wheels | 01:41 |
mordred | dstufft: because it's smart | 01:41 |
lifeless | dstufft: yes, but the pkg_resources stuff is broken by single-version-externally-managed | 01:41 |
lifeless | dstufft: and noone in cPython land is keen (AFAICT) on bringing up Requires or an equivalent into importlib. | 01:41 |
anteaya | mordred: so are there any operators that you can tell willing to learn to review requirements patches? or do they already? | 01:42 |
mordred | lifeless: in any case, this wasnt' so much a venue for us to give people feedback on how they could rethink their deployment decisions ... it was more about us listening to what they're doing, and what of our pending thoughts would make them have hives | 01:42 |
lifeless | dstufft: and since without being able to import multiple versions at once into the interpreter, its not at all equivalent to the ruby/java/js scenarios | 01:42 |
mordred | anteaya: I don't think they care what the requirements ARE | 01:42 |
dstufft | lifeless: yea, pkg_resources allows it, but pip explicitly disables it because the pkg_resources solution is super grody and breaks certain things :/ | 01:42 |
mordred | anteaya: they just one one contiguous set of them | 01:42 |
lifeless | mordred: yes, I get that | 01:42 |
anteaya | okay | 01:43 |
lifeless | mordred: I was more probing to see if it was an inconsistent cognitive thing | 01:43 |
anteaya | so good to know | 01:43 |
dstufft | lifeless: I was just mentioning that pkg_resources does allow it, and it was the default for awhile, but Python moved away from it towards venv because people grok'd that better | 01:43 |
dstufft | "python" being, "the community" | 01:43 |
lifeless | mordred: or whether we're super nice and they wish the other things supported this | 01:43 |
anteaya | mordred: what else came out of the day | 01:43 |
lifeless | mordred: or what | 01:43 |
lifeless | dstufft: aye, I know :) | 01:43 |
anteaya | in terms of stuff we need to ensure we keep doing? | 01:43 |
mordred | lifeless: nod | 01:44 |
nibalizer | mordred: hey you're online | 01:44 |
nibalizer | can you check the puppetforge credentials? | 01:44 |
nibalizer | jeblair: couldn't get them to work | 01:44 |
anteaya | is he still the only one with access? | 01:44 |
anteaya | mtreinish: on the etherpad you are not the only troll with the deep red | 01:45 |
anteaya | mtreinish: others are learning from you | 01:45 |
mtreinish | anteaya: heh, I didn't actually use etherpad all day, someone else has been using it :( | 01:46 |
nibalizer | anteaya: anyone with infra-root 'has access' the problem is the password mordred put in the special place doesn't seem to work | 01:46 |
sdague | anteaya: yeh, specifically when I asked "we're making you upgrade 150 libraries that you probably dont' need to because of find the common denominator" they all said "yup, not a problem" | 01:46 |
fungi | nibalizer: mordred: jeblair: i too tested the ones which got updated in hiera as of feb 24 (according to our git log there) and puppetforge told me to talk to the hand | 01:47 |
anteaya | mtreinish: ha ha ha | 01:47 |
*** rwsu has quit IRC | 01:47 | |
anteaya | sdague: wow | 01:47 |
anteaya | sdague: good piece of information to know | 01:47 |
sdague | yep | 01:47 |
mordred | it was very informative | 01:47 |
anteaya | no kidding | 01:48 |
anteaya | no way would I have guessed that would be the response | 01:48 |
jeblair | mordred, nibalizer: i'll just start password recovery tomorrow :) | 01:48 |
mordred | jeblair: cool | 01:48 |
morganfainberg | oh reading that etherpad is pretty awesome. | 01:48 |
*** yamahata has quit IRC | 01:49 | |
clarkb | sdague: its many more than 150 last I checked :) | 01:49 |
clarkb | sdague: first level is ~220 iirc | 01:49 |
sdague | clarkb: yeh, but a bunch are used only in single projects | 01:50 |
sdague | so 150 is probably a realistic estimate when you cross reference it with the services that people have actually deployed | 01:50 |
mordred | yup. they only build the thigns they need | 01:50 |
openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Refactor o_p::base into o_p::template https://review.openstack.org/162830 | 01:51 |
sdague | also, all the test libs are in that list | 01:51 |
mordred | nibalizer: dude. you're puppetting all the puppets | 01:51 |
clarkb | sdague: ya test libs are there. I just see the lines blending especially as we get oslo into more places | 01:51 |
mordred | nibalizer: btw - I spoke with $dude_from_puppetlabs | 01:51 |
clarkb | also debian will install all the things and buidl them so if reconsuming debian package stuff you likely get all of them | 01:51 |
nibalizer | hehe | 01:51 |
mordred | nibalizer: I'm sure he can't possibly like me | 01:51 |
mordred | nibalizer: I may have complained about the puppet2-puppet3 migration | 01:51 |
anteaya | mordred: nice job | 01:52 |
nibalizer | what did you say to him? | 01:52 |
nibalizer | hah! | 01:52 |
sdague | clarkb: mostly people are building their own packages it seems | 01:52 |
clarkb | sdague: competely from scratch and not just taking debians and build newer version? | 01:52 |
mordred | clarkb: unspecfied | 01:52 |
sdague | don't know | 01:52 |
mordred | but they're not just using what's in the distro | 01:52 |
clarkb | gotcha | 01:52 |
lifeless | see e.g. abnvil | 01:53 |
lifeless | bah | 01:53 |
lifeless | anvil | 01:53 |
lifeless | which does mass auto-building of python deps | 01:53 |
lifeless | harlowja_'s thing IIRC | 01:53 |
sdague | in the ovs talk in the morning they are specifically not using distro ovs because the distro is way behind | 01:53 |
harlowja_ | lifeless whatsss up | 01:53 |
harlowja_ | lol | 01:53 |
* harlowja_ didn't do it, lol | 01:53 | |
mordred | lifeless: I actually didnt' get the impression that building or maintaining packages was a pain point for anyone | 01:54 |
anteaya | sdague: interesting | 01:54 |
anteaya | so pain points | 01:54 |
anteaya | was their central agreement on the important 3? | 01:54 |
clarkb | mordred: probably helps to have 6 month cycles | 01:54 |
*** weshay has joined #openstack-infra | 01:54 | |
mordred | anteaya: pain points are mostly things like openstack not working well | 01:54 |
lifeless | mordred: sure, I was giving an example project to make sure its not | 01:54 |
morganfainberg | mordred, interesting. i've heard otherwise but come to think of it, it was all isolated and not complaining about building or maintaining packages, just that XXX package mngr sucks for Y reason. | 01:54 |
mordred | lifeless: ++ | 01:55 |
clarkb | mordred: so you aren't getting constant churn, just bug fixes then uplifts every once in a while | 01:55 |
mordred | clarkb: yup | 01:55 |
lifeless | mordred: which I know folk use, because my new deps on unittest2 got bug reports from anvil users | 01:55 |
mordred | lifeless: nice! | 01:55 |
morganfainberg | good ot know | 01:55 |
mordred | lifeless: I love learning of users that way | 01:55 |
anteaya | mordred: in the not working well metric, are we getting any better release to release or staying flat? | 01:55 |
lifeless | e.g. | 01:56 |
lifeless | https://github.com/testing-cabal/linecache2/issues/1 | 01:56 |
harlowja_ | the one from me doesn't count, lol | 01:56 |
* morganfainberg suddenly remembers there is a bug for testr that he needs to file/fix. | 01:56 | |
harlowja_ | i'm not exactly sure who benfilingbugs is, lol | 01:56 |
harlowja_ | odd name :-/ | 01:56 |
mordred | anteaya: didn't ask in that framework - one of the lines of questioning led to "we need a better mechanism for you to tell us what you'd like to run in the future but aren't now and why" | 01:56 |
clarkb | mordred: to run being deps? | 01:57 |
lifeless | harlowja_: https://github.com/BenFilingBugs - joined 9th march :) | 01:57 |
anteaya | I'd like somethig along the lines of prioritze the top 5 things you would like to help openstack work better | 01:57 |
harlowja_ | ya, its not an alias for me, i swear lifeless | 01:57 |
harlowja_ | lol | 01:57 |
mordred | clarkb: no - services | 01:57 |
mordred | clarkb: like, if was ask "what are you running" | 01:58 |
anteaya | I'm not interested in more, I'm interested in making what we have work | 01:58 |
mordred | clarkb: there is a follow up question which is "what aren't you running that you wish you were" | 01:58 |
clarkb | mordred: gotcha | 01:58 |
mordred | designate came up in that context | 01:58 |
mordred | many people want to run it | 01:58 |
mordred | but aren't beacuse there area couple of blocker things or something | 01:58 |
clarkb | mordred: I asked about deps because I think thats one of the tricky things to solve if/when we have our own packages | 01:58 |
clarkb | mordred: its not building them its what to build | 01:58 |
mordred | clarkb: ++ | 01:59 |
*** esker has joined #openstack-infra | 02:00 | |
*** esker has quit IRC | 02:00 | |
*** armax has quit IRC | 02:00 | |
*** esker has joined #openstack-infra | 02:01 | |
*** craigbr has joined #openstack-infra | 02:01 | |
harlowja_ | so i can tell u guys how anvil works, if u care :-P | 02:03 |
harlowja_ | not that much of a secret, ha | 02:03 |
*** mriedem has quit IRC | 02:03 | |
anteaya | since I'm not a guy, I'll move along | 02:03 |
harlowja_ | *gals/guys | 02:03 |
lifeless | mordred: do you think the ops would be ok with e.g. https://github.com/grpc/grpc/tree/master as a dep ? | 02:03 |
harlowja_ | human beings, lol | 02:04 |
*** craigbr has quit IRC | 02:04 | |
greghaynes | grpc :) | 02:04 |
mordred | lifeless: I do not think they would care at all | 02:05 |
harlowja_ | my guess is if the shit works, they'd accept nearly anything, lol | 02:05 |
mordred | yup | 02:05 |
mordred | greghaynes: did you see waynr earlier today ping jhesketh about PR support for zuul | 02:06 |
mordred | ? | 02:06 |
mordred | greghaynes: waynr said something about starting to work on it | 02:06 |
*** craigbr has joined #openstack-infra | 02:06 | |
lifeless | greghaynes: yes, on a shallow looks its pretty much the ideal RPC layer I've had in my head for a year or two | 02:07 |
lifeless | greghaynes: minor insubstantial tweaks like protobuf instead of capnproto | 02:07 |
jeblair | mordred: i suggested greghaynes and waynr work on a spec. they agreed that was a good idea. :) | 02:07 |
mordred | jeblair: awesome | 02:07 |
greghaynes | mordred: yep, we chatted | 02:08 |
mordred | greghaynes, jeblair: I just happened to see the mention on my phone briefly, but wasn't in a spot to respond ;) | 02:08 |
greghaynes | lifeless: Yea, I havent dived deep into it but I hear good things | 02:08 |
sdague | lifeless: so the python readme for that says 'pre Alpha, usable with limitations' | 02:08 |
sdague | which seems... concerning | 02:08 |
dstufft | lifeless: newer pips hide the output from compiling bytecode fwiw | 02:08 |
dstufft | even newer pips (develop branch) more or less hide all output from the setup.py unless there's an error | 02:09 |
lifeless | dstufft: 6.0.7. seems to show this AFAICT. | 02:09 |
lifeless | dstufft: nice | 02:09 |
jhesketh | mordred: oh, I don't think I saw that... that sounds cool | 02:09 |
lifeless | sdague: possibly yes, obviously due diligence needed. But we have lots of argh my eyes in our rpc, and I think its actually fairly high up the list in terms of sources of odd behaviours | 02:09 |
dstufft | lifeless: hm, maybe we regressed it at some point, 1.5 I think was the version that silenced the syntax errors | 02:10 |
lifeless | sdague: not to mention the headaches with HA etc etc today | 02:10 |
dstufft | (or mayb eit only silenced it on certain versions of python, I forget) | 02:10 |
anteaya | mordred: so if you can manage to get any disernable amount of operators to agree what the top three things they would like to see to make openstack work better, I would like to hear that list | 02:10 |
dstufft | https://github.com/pypa/pip/pull/2487 new output! | 02:10 |
*** ivar-lazzaro has quit IRC | 02:11 | |
jhesketh | waynr: ping | 02:11 |
*** unicell has quit IRC | 02:12 | |
greghaynes | jhesketh: Hey, I left another round of comments on your connection patch, I think you must have overlooked my non-inline original comment on removing lib/gerrit.py? | 02:12 |
lifeless | I think I'm going to have to do a spike on it though | 02:12 |
greghaynes | jhesketh: Sorry to keep making you circle back on it too btw, its just a big change so theres a lot to process :) | 02:13 |
jhesketh | greghaynes: hey, yep, I saw that. Sorry I didn't get a chance to update it | 02:13 |
jhesketh | greghaynes: oh not at all, I really appreciate your reviews and attention to detail | 02:13 |
greghaynes | its np, just wanted to make sure you saw it | 02:13 |
jhesketh | it's a big change that has been sitting there for a long time so I need to refresh some of my state too :-) | 02:13 |
*** craigbr has quit IRC | 02:14 | |
*** bhuvan has quit IRC | 02:14 | |
*** ghostpl_ has quit IRC | 02:17 | |
*** craigbr has joined #openstack-infra | 02:19 | |
*** salv-orlando has quit IRC | 02:20 | |
*** craigbr has quit IRC | 02:21 | |
*** craigbr has joined #openstack-infra | 02:22 | |
*** craigbr has quit IRC | 02:26 | |
*** baoli has joined #openstack-infra | 02:32 | |
*** coolsvap has joined #openstack-infra | 02:33 | |
clarkb | nibalizer: crinkle how do I query if puppet agent is disabled without running a --noop run? | 02:34 |
clarkb | nibalizer: googling suggests puppet status may do it but the suggested -S flag does not exist in our install | 02:34 |
mordred | clarkb: look for the lockfile? | 02:35 |
*** patrickeast has quit IRC | 02:35 | |
*** amotoki has quit IRC | 02:35 | |
clarkb | mordred: I think that is platform dependent though (I am trying to fix your ansible change) | 02:36 |
nibalizer | ya i would [ -f `puppet agent --configprint $nameoflockfile` ] | 02:36 |
nibalizer | but im not really reccomending you do that | 02:36 |
clarkb | nibalizer: thanks | 02:36 |
clarkb | nibalizer: what would you recommend? | 02:36 |
mordred | clarkb: oh - btw - I had this kinda-hacky idea I'm going to test next time I get a sec of how to deal with disabling things when we move to ansible->puppet apply | 02:36 |
mordred | clarkb: which is, we'll likely have an inventory directory anyway - which is a thing that lets you have multiple inventories - we could drop a "Disabled" file in there that has a [disabled] group in it - and we can run puppet on '*;!disabled' | 02:37 |
mordred | clarkb: also - I'm actually online - can I help fix the patch? | 02:37 |
clarkb | mordred: sure, I think I have it with nibalizers suggestion though :) | 02:38 |
clarkb | mordred: so up to you if I should not write it and have you do it instead | 02:38 |
mordred | clarkb: awesome | 02:38 |
mordred | clarkb: nope - go for it | 02:38 |
nibalizer | clarkb: yea for scripting i would do that | 02:38 |
nibalizer | i thought you were on an interactive shell | 02:38 |
*** Ryan_Lane has joined #openstack-infra | 02:42 | |
openstackgerrit | Clark Boylan proposed openstack-infra/ansible-puppet: Add ability to run puppet with passed-in data https://review.openstack.org/159159 | 02:43 |
clarkb | mordred: ^ | 02:43 |
mordred | clarkb: woot | 02:44 |
*** craigbr has joined #openstack-infra | 02:44 | |
*** coolsvap has quit IRC | 02:45 | |
*** coolsvap has joined #openstack-infra | 02:46 | |
mordred | clarkb: looks great - one small -1 | 02:46 |
clarkb | mordred: so now we just need my puppet-project_config change and a yet to be written system-config change to fetch project-config's HEAD during our playbooks and pass that in as a fact | 02:46 |
clarkb | mordred: looking | 02:46 |
*** coolsvap has quit IRC | 02:46 | |
*** weshay has quit IRC | 02:46 | |
*** coolsvap has joined #openstack-infra | 02:46 | |
mordred | clarkb: yes | 02:47 |
*** achanda has joined #openstack-infra | 02:47 | |
clarkb | mordred: can I pass that to fail_json or do I need to use exit_json? | 02:49 |
clarkb | mordred: I would look up the docs but I know I won't get an answer :P | 02:49 |
mordred | clarkb: you can pass to fail_json | 02:50 |
mordred | clarkb: oh! wait | 02:50 |
*** woodster_ has quit IRC | 02:50 | |
mordred | I do the other thing below | 02:50 |
mordred | look at lines 166-168 | 02:50 |
clarkb | right | 02:50 |
clarkb | that is what prompted my question | 02:50 |
clarkb | I did just check the docs btw | 02:51 |
clarkb | not useful | 02:51 |
mordred | I will learn the answer | 02:51 |
clarkb | (really they just need autodoc | 02:51 |
mordred | clarkb: where did you check the docs? | 02:51 |
*** unicell has joined #openstack-infra | 02:51 | |
clarkb | mordred: http://docs.ansible.com/developing_modules.html | 02:51 |
mordred | awesome | 02:51 |
*** markvoelker has joined #openstack-infra | 02:51 | |
mordred | clarkb: exit_json and fail_json have the same interface | 02:52 |
*** hdd_ has quit IRC | 02:52 | |
mordred | clarkb: the difference seems to be that fail_json injects a "failed=True" into the kwargs | 02:52 |
clarkb | thanks I will just use fail_json as it avoid needing to ^ | 02:52 |
*** craigbr has quit IRC | 02:52 | |
clarkb | should I set changed = False? | 02:53 |
mordred | clarkb: and exit_json has "if not 'changed' in kwargs: kwargs['changed'] = False" | 02:53 |
*** craigbr has joined #openstack-infra | 02:53 | |
clarkb | perfect | 02:53 |
mordred | clarkb: but I agree- that should be documented betterer | 02:53 |
openstackgerrit | Clark Boylan proposed openstack-infra/ansible-puppet: Add ability to run puppet with passed-in data https://review.openstack.org/159159 | 02:53 |
clarkb | mordred: did you abandon the system-config change to pass in the HEAD value? I thought you had written one but I can't find it | 02:54 |
clarkb | also dinner now will check back after | 02:54 |
*** craigbr has quit IRC | 02:54 | |
mordred | clarkb: +2 | 02:55 |
mordred | clarkb: I thought I'd written one too ... but maybe we're both just drunk | 02:55 |
*** michchap has joined #openstack-infra | 03:01 | |
*** wenlock has quit IRC | 03:05 | |
clarkb | mordred: OH you updated my change | 03:11 |
mordred | clarkb: I did? | 03:11 |
clarkb | ya, I will clean it up given the feedback and my puppet-project_config change and add a depends-on | 03:11 |
*** Ryan_Lane has quit IRC | 03:12 | |
clarkb | arg I like spamaps suggestion but now I have to go rtfs | 03:14 |
*** ddieterl_ has quit IRC | 03:16 | |
*** ghostpl_ has joined #openstack-infra | 03:18 | |
*** SumitNaiksatam has quit IRC | 03:18 | |
*** Ryan_Lane has joined #openstack-infra | 03:18 | |
clarkb | mordred: https://github.com/ansible/ansible-modules-core/blob/devel/source_control/git.py#L684-L687 we just need to get them to put that info in docstrings instead of comments :) | 03:19 |
mordred | clarkb: what's the SpamapS suggestion? | 03:21 |
*** salv-orlando has joined #openstack-infra | 03:21 | |
*** bhuvan has joined #openstack-infra | 03:22 | |
clarkb | mordred: to use the ansible git module to do the ls-remote for us | 03:23 |
*** sdake__ has quit IRC | 03:23 | |
clarkb | mordred: will push change up in just a moment | 03:23 |
*** jyuso1 has quit IRC | 03:24 | |
mordred | kk | 03:24 |
openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Pass in a ref for the project_config repo https://review.openstack.org/154305 | 03:25 |
clarkb | mordred: ^ | 03:25 |
clarkb | it losely depends on https://review.openstack.org/#/c/154300/2 too but we can merge in either order so I didn't add that one just to simplify | 03:25 |
*** salv-orlando has quit IRC | 03:26 | |
mordred | clarkb: looks great! | 03:26 |
*** ghostpl_ has quit IRC | 03:28 | |
clarkb | cool we are close to removing that race | 03:28 |
clarkb | mordred: any chance you can review 154300 as well? | 03:31 |
*** dims_ has quit IRC | 03:32 | |
*** otter768 has quit IRC | 03:32 | |
*** esker has quit IRC | 03:33 | |
*** ujuc has joined #openstack-infra | 03:33 | |
mordred | clarkb: done. | 03:33 |
*** otter768 has joined #openstack-infra | 03:34 | |
nibalizer | so.... puppet master in jruby? aka puppet-server | 03:34 |
nibalizer | thoughts? | 03:34 |
mordred | nibalizer: zero thoughts | 03:34 |
nibalizer | im seriously cinsideringusing it in gozer land | 03:34 |
clarkb | nibalizer: we are trying to remove master :) | 03:34 |
mordred | nibalizer: I want the master to go away | 03:34 |
mordred | very close to being ready for that | 03:34 |
*** craigbr has joined #openstack-infra | 03:34 | |
nibalizer | hrmmm | 03:35 |
nibalizer | i dont think youve solved the hiera problem | 03:35 |
mordred | why not/ | 03:35 |
mordred | ? | 03:35 |
nibalizer | how did you solve it | 03:35 |
mordred | I mean, I'm not done solving it - but ansible will pass values in to the server so that they're available to puppet apply | 03:36 |
mordred | I Think we're exceptionally close though - maybe one more week of me not being at a conference | 03:36 |
mordred | honestly | 03:36 |
clarkb | mordred: yup and that change I just updated is a major step in making that happen | 03:37 |
mordred | yup | 03:37 |
*** claudiub has quit IRC | 03:37 | |
mordred | but - long day - I'm tired - AFKs | 03:38 |
clarkb | nibalizer: see https://review.openstack.org/159159 | 03:38 |
clarkb | mordred: thank you for the reviews | 03:39 |
clarkb | I will pester other people in the morning and hope to fix this problem tomorrow | 03:39 |
*** achanda has quit IRC | 03:40 | |
*** otter768 has quit IRC | 03:43 | |
*** tiswanso has quit IRC | 03:43 | |
*** ayoung has quit IRC | 03:44 | |
nibalizer | mordred: idebtifying which values to look up and pass in... how to solve that | 03:44 |
*** koolhead17 has joined #openstack-infra | 03:45 | |
clarkb | nibalizer: its the same as today, you write them down in yaml somewhere then use them | 03:45 |
*** hdd_ has joined #openstack-infra | 03:46 | |
nibalizer | clarkb: you need a mappibg of some kind from secrets to nodes though | 03:47 |
*** ddieterly has joined #openstack-infra | 03:47 | |
clarkb | nibalizer: ya thats ansible | 03:47 |
clarkb | nibalizer: see https://review.openstack.org/154305 | 03:47 |
*** sdake_ has joined #openstack-infra | 03:47 | |
nibalizer | ah | 03:49 |
nibalizer | so all the secrets are just gonna go in ansible? | 03:49 |
clarkb | I don't think we hae solved that yet. getting ansible to use the hiera data is probably pretty easy | 03:49 |
nibalizer | hostvars or wgatever they yse that isbt hiera? | 03:49 |
clarkb | so we might expose something like hostvars.localhost.hiera | 03:50 |
clarkb | I dunno if mordred has decided on something to start with there | 03:50 |
openstackgerrit | Brian Hunter proposed openstack-infra/nodepool: Pretty Age in nodepool list https://review.openstack.org/161994 | 03:50 |
clarkb | but then you can give subtrees of that to each node as necessary | 03:51 |
*** camunoz has quit IRC | 03:51 | |
*** ddieterly has quit IRC | 03:51 | |
*** dboik_ has quit IRC | 03:51 | |
nibalizer | cool okay that was the missing piece | 03:52 |
*** markvoelker has quit IRC | 03:53 | |
*** markvoelker has joined #openstack-infra | 03:53 | |
*** Ryan_Lane has quit IRC | 03:56 | |
*** camunoz has joined #openstack-infra | 03:56 | |
*** markvoelker has quit IRC | 03:57 | |
*** craigbr has quit IRC | 03:58 | |
*** sputnik13 has joined #openstack-infra | 03:59 | |
*** Longgeek has joined #openstack-infra | 03:59 | |
clarkb | btw https://etherpad-dev.openstack.org/p/clarkb-testing I restarted the etherpad service on etherpad-dev | 03:59 |
clarkb | the git checkout was for the latest version and the robots.txt was rewritten so jesusaurus' changes made things happier | 03:59 |
clarkb | the update seemed to take a bit of time (db migrations I guess?) we probably want to wait for after the ops summit to upgrade | 04:00 |
*** vilobhmm has joined #openstack-infra | 04:00 | |
clarkb | anyways please to test | 04:00 |
*** salv-orlando has joined #openstack-infra | 04:01 | |
*** sputnik13 has quit IRC | 04:02 | |
*** sabeen has joined #openstack-infra | 04:02 | |
*** achanda has joined #openstack-infra | 04:06 | |
openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: New job for in-tree lbaasv2 tempest tests https://review.openstack.org/161432 | 04:08 |
jesusaurus | clarkb: how many local changes are in the vcsrepo? im just curious how many files are being recreated with every puppet run | 04:08 |
clarkb | jesusaurus: what do you mean by being recreated with every puppet run? its one change in one file of the provider iirc to be more git like | 04:09 |
*** garyh has quit IRC | 04:09 | |
jesusaurus | well if there's local change to robots.txt, then every puppet run will essentially reset --hard to HEAD and then reapply the local change | 04:10 |
clarkb | oh you mean local changes in the repo itself not vcsrepo, I misread | 04:10 |
clarkb | ya its just the robots.txt | 04:10 |
jesusaurus | cool | 04:11 |
clarkb | 3 line diff iirc | 04:11 |
clarkb | because we host from / not /p/ | 04:11 |
jesusaurus | oh right | 04:11 |
clarkb | for historical reasons (keeps compat with OLD urls) | 04:11 |
jesusaurus | its a little annoying that the change will be undone and redone every puppet run, but its not a huge deal | 04:12 |
clarkb | we should maybe try a PR to make robots.txt in the repo a thing you update like the config? | 04:13 |
clarkb | or make robots.txt location configurable | 04:13 |
clarkb | then point to a different file | 04:13 |
*** praveenkumar has joined #openstack-infra | 04:13 | |
*** craigbr has joined #openstack-infra | 04:14 | |
*** salv-orlando has quit IRC | 04:14 | |
jesusaurus | i like making robots.txt something you can update like the config, but i dont have the source in front of me | 04:14 |
openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: neutron-*aas pylint jobs moving into pep8 job https://review.openstack.org/160915 | 04:21 |
dougwig | if anyone from infra core is around, i'd appreciate a review on this one, so i can start iterating on our in-tree tempest hooks: https://review.openstack.org/#/c/161432/ | 04:23 |
*** camunoz has quit IRC | 04:23 | |
*** markvoelker has joined #openstack-infra | 04:24 | |
*** camunoz has joined #openstack-infra | 04:24 | |
*** Ryan_Lane has joined #openstack-infra | 04:24 | |
*** craigbr has quit IRC | 04:28 | |
*** craigbr has joined #openstack-infra | 04:28 | |
*** amotoki has joined #openstack-infra | 04:30 | |
*** Daisy_ has quit IRC | 04:31 | |
clarkb | dougwig: done | 04:31 |
dougwig | clarkb: thank you | 04:32 |
*** dims has joined #openstack-infra | 04:32 | |
*** BharatK has joined #openstack-infra | 04:33 | |
openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Test that dib failures don't stop node builds https://review.openstack.org/162237 | 04:37 |
openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Accomodate ready nodes in min-ready checks https://review.openstack.org/162714 | 04:37 |
openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Don't allocate nodes to providers without images https://review.openstack.org/162255 | 04:37 |
*** Somay has joined #openstack-infra | 04:37 | |
openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: New job for in-tree lbaasv2 tempest tests https://review.openstack.org/161432 | 04:37 |
dougwig | what does dsvm stand for, btw? | 04:38 |
*** dims has quit IRC | 04:38 | |
StevenK | devstack virtual machine, I think | 04:38 |
*** Ryan_Lane has quit IRC | 04:38 | |
dougwig | ahh. ty. | 04:38 |
openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Test that dib failures don't stop node builds https://review.openstack.org/162237 | 04:39 |
openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Accomodate ready nodes in min-ready checks https://review.openstack.org/162714 | 04:39 |
openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Don't allocate nodes to providers without images https://review.openstack.org/162255 | 04:39 |
clarkb | ianw: ^ hows that look? | 04:39 |
*** yfried|afk is now known as yfried | 04:40 | |
*** hughhalf is now known as hugh_afk | 04:40 | |
*** dims has joined #openstack-infra | 04:40 | |
ianw | clarkb: ok, lgtm :) i think it reads clearer? | 04:40 |
ianw | clarkb: really, we should tell the allocator "i need X and would like Y to be ready, if you can" | 04:41 |
*** markvoelker has quit IRC | 04:41 | |
clarkb | ianw: ya | 04:42 |
clarkb | dougwig: one more set of comments, but its just about ready | 04:44 |
dougwig | clarkb: ahh, thanks. i should've thought of that. | 04:44 |
*** dims has quit IRC | 04:45 | |
clarkb | jenkins02 has 4k threads, look for it to be next in line of thread leak | 04:46 |
*** baoli has quit IRC | 04:48 | |
*** alexpilotti has joined #openstack-infra | 04:48 | |
*** ddieterly has joined #openstack-infra | 04:48 | |
openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: New job for in-tree lbaasv2 tempest tests https://review.openstack.org/161432 | 04:48 |
clarkb | zaro: still awake? we never converted the gerrit db to utf8 right? | 04:49 |
clarkb | zaro: there is a thread on the infra list asking about korean characters in review comments and aiui those won't work correct with a non utf8 db right? | 04:49 |
*** baoli has joined #openstack-infra | 04:50 | |
ianw | clarkb: that apt-get tty thing is a bit of a pickle. https://review.openstack.org/#/c/160480/ . strace shows it dying with EPIPE which suggests to me whatever it was outputting to disappeared on it | 04:52 |
*** ddieterly has quit IRC | 04:53 | |
ianw | maybe the devstack python logger thing ... depending on the setup | 04:53 |
clarkb | zaro: nevermind I dug up the info from the databases information schema db | 04:53 |
clarkb | ianw: huh | 04:54 |
*** yfried is now known as yfried|afk | 04:56 | |
clarkb | ianw: I think that may mean dpkg was requesting the input and died because it couldn't get it? | 04:56 |
*** baoli has quit IRC | 04:56 | |
clarkb | ianw: then SIGPIPE went to the parent and everything else in the group just gets the SIGTTIN? | 04:57 |
*** Somay has quit IRC | 04:58 | |
*** yfried|afk is now known as yfried | 04:58 | |
clarkb | oh wait I see :) its late for me I should stop trying to think so hard | 04:58 |
ianw | clarkb: don't think so, dpkg dies | 04:58 |
*** kiseok7_ has quit IRC | 04:59 | |
clarkb | ya if the python logger thing explodes that may cause it to all fall apart | 04:59 |
ianw | clarkb: just like say "yes | head -1" or something where the output goes away on you | 04:59 |
clarkb | I am not familiar with the python logger though would need to read up on it | 04:59 |
*** Somay has joined #openstack-infra | 04:59 | |
clarkb | but this gives me more info to debug I can respin a test env to run apt-get install | outputfilter interactively to see if I can get it to explode | 05:01 |
ianw | clarkb: yes step away from irc :) but yeah, it's so many levels deep by that point who really knows ... | 05:01 |
clarkb | I have so many things to do, trying to catch up after spending a large part of the day afk | 05:02 |
clarkb | but I think I should call it a day, good night | 05:02 |
*** baoli has joined #openstack-infra | 05:02 | |
*** craigbr has quit IRC | 05:04 | |
*** baoli has quit IRC | 05:07 | |
*** hdd_ has quit IRC | 05:08 | |
*** dhritishikhar has joined #openstack-infra | 05:08 | |
*** Somay_ has joined #openstack-infra | 05:08 | |
*** Somay has quit IRC | 05:09 | |
*** yfried is now known as yfried|afk | 05:10 | |
*** Sukhdev has joined #openstack-infra | 05:11 | |
*** yfried|afk is now known as yfried | 05:12 | |
*** ghostpl_ has joined #openstack-infra | 05:13 | |
*** salv-orlando has joined #openstack-infra | 05:13 | |
*** Daisy_ has joined #openstack-infra | 05:13 | |
Sukhdev | I am seeing the gate failure - can anybody provide any insight - see here - https://review.openstack.org/#/c/162778/ | 05:15 |
*** Alexandra_ has quit IRC | 05:15 | |
*** unicell1 has joined #openstack-infra | 05:18 | |
*** unicell has quit IRC | 05:18 | |
*** salv-orlando has quit IRC | 05:19 | |
*** ghostpl_ has quit IRC | 05:20 | |
*** salv-orlando has joined #openstack-infra | 05:22 | |
openstackgerrit | Steve Leon proposed openstack-infra/project-config: Adding new project: cue-dashboard to stackforge https://review.openstack.org/162804 | 05:23 |
*** salv-orlando has quit IRC | 05:26 | |
*** yamamoto has quit IRC | 05:27 | |
*** yfried has quit IRC | 05:27 | |
*** craigbr has joined #openstack-infra | 05:30 | |
*** stevemar has quit IRC | 05:31 | |
*** markvoelker has joined #openstack-infra | 05:38 | |
*** hdd_ has joined #openstack-infra | 05:38 | |
*** otter768 has joined #openstack-infra | 05:44 | |
openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Refactor o_p::base into o_p::template https://review.openstack.org/162830 | 05:45 |
*** ildikov has quit IRC | 05:47 | |
*** craigbr has quit IRC | 05:48 | |
*** otter768 has quit IRC | 05:49 | |
*** ddieterly has joined #openstack-infra | 05:49 | |
*** sputnik13 has joined #openstack-infra | 05:49 | |
*** ddieterly has quit IRC | 05:53 | |
*** sputnik13 has quit IRC | 05:54 | |
*** achanda has quit IRC | 05:54 | |
*** achanda has joined #openstack-infra | 05:55 | |
*** dhritishikhar has quit IRC | 05:55 | |
*** kiseok7_ has joined #openstack-infra | 05:56 | |
*** harlowja_ is now known as harlowja_away | 05:57 | |
*** Somay_ has quit IRC | 05:57 | |
openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Refactor o_p::base into o_p::template https://review.openstack.org/162830 | 05:59 |
*** mrmartin has joined #openstack-infra | 06:01 | |
*** yamamoto has joined #openstack-infra | 06:02 | |
*** Longgeek has quit IRC | 06:04 | |
*** BharatK has quit IRC | 06:06 | |
*** salv-orlando has joined #openstack-infra | 06:08 | |
*** salv-orlando has quit IRC | 06:08 | |
*** achanda has quit IRC | 06:11 | |
*** achanda has joined #openstack-infra | 06:11 | |
*** craigbr has joined #openstack-infra | 06:15 | |
*** BharatK has joined #openstack-infra | 06:20 | |
*** dims has joined #openstack-infra | 06:21 | |
*** vilobhmm1 has joined #openstack-infra | 06:22 | |
*** Longgeek has joined #openstack-infra | 06:23 | |
*** vilobhmm has quit IRC | 06:23 | |
*** dims has quit IRC | 06:26 | |
*** achanda has quit IRC | 06:28 | |
*** salv-orlando has joined #openstack-infra | 06:28 | |
*** sgordon has quit IRC | 06:30 | |
*** david-lyle has quit IRC | 06:30 | |
*** mrmartin has quit IRC | 06:30 | |
*** yfried has joined #openstack-infra | 06:36 | |
*** hdd_ has quit IRC | 06:38 | |
*** salv-orlando has quit IRC | 06:38 | |
*** david-lyle has joined #openstack-infra | 06:39 | |
*** sgordon has joined #openstack-infra | 06:40 | |
*** yfried_ has joined #openstack-infra | 06:40 | |
*** ildikov has joined #openstack-infra | 06:40 | |
*** yfried has quit IRC | 06:41 | |
*** yfried_ is now known as yfried|afk | 06:44 | |
*** coolsvap is now known as coolsvap|afk | 06:47 | |
*** mrmartin has joined #openstack-infra | 06:48 | |
*** ddieterly has joined #openstack-infra | 06:50 | |
openstackgerrit | yolanda.robla proposed openstack-infra/puppet-zuul: Add zuul_override and fetch_url parameters to zuul.conf https://review.openstack.org/162591 | 06:52 |
*** vilobhmm1 has quit IRC | 06:53 | |
*** vilobhmm has joined #openstack-infra | 06:54 | |
*** ddieterly has quit IRC | 06:54 | |
*** salv-orlando has joined #openstack-infra | 06:54 | |
*** coolsvap|afk is now known as coolsvap | 06:58 | |
*** salv-orlando has quit IRC | 06:59 | |
*** dhritishikhar has joined #openstack-infra | 07:01 | |
*** ghostpl_ has joined #openstack-infra | 07:01 | |
*** yfried|afk is now known as yfried_ | 07:04 | |
*** garyk has joined #openstack-infra | 07:06 | |
*** ghostpl_ has quit IRC | 07:09 | |
*** Sukhdev has quit IRC | 07:10 | |
*** david-lyle is now known as david-lyle_afk | 07:11 | |
*** AJaeger has joined #openstack-infra | 07:12 | |
*** AJaeger has joined #openstack-infra | 07:12 | |
AJaeger | jeblair, the propose-project-config-update job failed - see https://jenkins.openstack.org/view/All/job/propose-project-config-update/1/console | 07:13 |
*** che-arne has quit IRC | 07:15 | |
*** tnovacik has joined #openstack-infra | 07:19 | |
*** mrunge has joined #openstack-infra | 07:19 | |
openstackgerrit | Merged openstack-infra/project-config: Adding xstatic notifications to openstack-horizon https://review.openstack.org/162174 | 07:20 |
openstackgerrit | yolanda.robla proposed openstack-infra/puppet-zuul: Add zuul_override and fetch_url parameters to zuul.conf https://review.openstack.org/162591 | 07:21 |
*** scheuran has joined #openstack-infra | 07:21 | |
*** chlong has quit IRC | 07:21 | |
openstackgerrit | yolanda.robla proposed openstack-infra/system-config: Add configurable url for OpenStack status https://review.openstack.org/162169 | 07:23 |
*** nithyag__ has quit IRC | 07:24 | |
*** nithyag__ has joined #openstack-infra | 07:25 | |
*** jcoufal has joined #openstack-infra | 07:25 | |
*** Daisy__ has joined #openstack-infra | 07:26 | |
*** Daisy_ has quit IRC | 07:27 | |
*** jistr has joined #openstack-infra | 07:31 | |
*** mpaolino has joined #openstack-infra | 07:32 | |
*** tkelsey has joined #openstack-infra | 07:33 | |
*** openstackgerrit has quit IRC | 07:35 | |
*** koolhead17 has quit IRC | 07:35 | |
*** openstackgerrit has joined #openstack-infra | 07:35 | |
*** tkelsey has quit IRC | 07:38 | |
*** BharatK has quit IRC | 07:39 | |
*** tkelsey has joined #openstack-infra | 07:39 | |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add initial compute functional tests to Shade https://review.openstack.org/158463 | 07:42 |
*** koolhead17 has joined #openstack-infra | 07:43 | |
*** HeOS has joined #openstack-infra | 07:44 | |
*** gampel has joined #openstack-infra | 07:44 | |
*** otter768 has joined #openstack-infra | 07:45 | |
*** otter768 has quit IRC | 07:50 | |
*** ddieterly has joined #openstack-infra | 07:51 | |
*** BharatK has joined #openstack-infra | 07:52 | |
*** coolsvap is now known as coolsvap|afk | 07:53 | |
*** sabeen has quit IRC | 07:55 | |
*** ddieterly has quit IRC | 07:55 | |
openstackgerrit | Yanis Guenane proposed openstack-infra/project-config: New project: puppet-modulesync-configs https://review.openstack.org/162902 | 07:55 |
spredzy | Morning all, just submitted the creation of the puppet-modulesync-configs project https://review.openstack.org/#/c/162902/ | 07:57 |
*** andreykurilin_ has joined #openstack-infra | 07:57 | |
*** ociuhandu has joined #openstack-infra | 07:59 | |
*** coolsvap|afk is now known as coolsvap | 08:00 | |
*** sergsh has joined #openstack-infra | 08:01 | |
anteaya | Sukhev I think you saw _a_ gate failure: http://logs.openstack.org/78/162778/2/gate/gate-tempest-dsvm-neutron-full/1fe5208/console.html#_2015-03-10_02_25_52_073 looks like a time out for some reason | 08:01 |
*** markus_z has joined #openstack-infra | 08:03 | |
*** jyuso1 has joined #openstack-infra | 08:05 | |
*** achuprin_ has quit IRC | 08:05 | |
*** MaxV has joined #openstack-infra | 08:07 | |
*** rcarrillocruz has joined #openstack-infra | 08:07 | |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add initial compute functional tests to Shade https://review.openstack.org/158463 | 08:08 |
*** ghostpl_ has joined #openstack-infra | 08:10 | |
*** prmtl has joined #openstack-infra | 08:12 | |
*** dhritishikhar has quit IRC | 08:12 | |
*** Hal has joined #openstack-infra | 08:13 | |
*** Hal is now known as Guest88753 | 08:13 | |
*** liusheng has joined #openstack-infra | 08:14 | |
*** ghostpl_ has quit IRC | 08:15 | |
*** yamahata has joined #openstack-infra | 08:18 | |
*** achuprin_ has joined #openstack-infra | 08:18 | |
*** dhritishikhar has joined #openstack-infra | 08:18 | |
*** luqas has joined #openstack-infra | 08:19 | |
GheRivero | morning all | 08:22 |
*** claudiub has joined #openstack-infra | 08:25 | |
openstackgerrit | Thomas Morin proposed openstack-infra/project-config: Add networking-bagpipe-l2 project https://review.openstack.org/161290 | 08:29 |
*** andreykurilin_ has quit IRC | 08:29 | |
gampel | Hi The zull gate merge job is not kicking in our project after review, is there a problem is there a problem in the infrastructure ? https://review.openstack.org/#/c/162558/ | 08:31 |
*** yfried_ is now known as yfried|afk | 08:36 | |
*** jlanoux has joined #openstack-infra | 08:38 | |
anteaya | gampel: you set workflow +1 before you set code review +2: https://review.openstack.org/#/c/162558/ | 08:40 |
anteaya | try removing workflow +1 and then adding it again | 08:40 |
anteaya | morning GheRivero | 08:41 |
*** achuprin_ has quit IRC | 08:41 | |
*** mpavone_ has joined #openstack-infra | 08:41 | |
*** salv-orlando has joined #openstack-infra | 08:42 | |
*** hashar has joined #openstack-infra | 08:42 | |
*** salv-orlando has quit IRC | 08:46 | |
gampel | thank you anteaya it is working now | 08:47 |
anteaya | great, you're welcome | 08:47 |
anteaya | in future you don't need to set code review +1 before you set code review +2 | 08:47 |
anteaya | but you do need code review +2 either before or at the same time as workflow +1 | 08:48 |
AJaeger | anteaya, great detective work! | 08:48 |
anteaya | AJaeger: hey there | 08:48 |
anteaya | I had some time between meetings | 08:48 |
AJaeger | ;) | 08:48 |
anteaya | now I understand how you can get so much work done | 08:49 |
anteaya | it is a ghost town in here at this time of day | 08:49 |
anteaya | and thanks | 08:49 |
markus_z | What can I do when 2 reviews (1x abandondend, 1x ok) have the same change-id and I have a third patch which needs the "depends-on" keyword to reference the "ok"-review? | 08:50 |
markus_z | See "depends-on" in https://review.openstack.org/#/c/162678/ | 08:51 |
*** ddieterly has joined #openstack-infra | 08:51 | |
*** alexpilotti_ has joined #openstack-infra | 08:52 | |
*** jerryz has quit IRC | 08:52 | |
*** arxcruz has joined #openstack-infra | 08:52 | |
*** dhritishikhar has quit IRC | 08:53 | |
anteaya | this came up I think last week | 08:53 |
anteaya | the overall story starts with using the same change id wasn't meant to track patches across projects | 08:54 |
*** alexpilotti has quit IRC | 08:54 | |
*** alexpilotti_ is now known as alexpilotti | 08:54 | |
anteaya | it was meant to track the _same_ patch across branches | 08:54 |
anteaya | such as backporting a fix | 08:54 |
AJaeger | markus_z, either don't use depends-on or change the patches so that it does not depend on an abandoned one | 08:54 |
AJaeger | anteaya, that's not my understanding, it was meant for crossing projects | 08:54 |
anteaya | AJaeger: oh | 08:55 |
*** yfried|afk is now known as yfried_ | 08:55 | |
AJaeger | sorry, meeting time - I leave... | 08:55 |
AJaeger | anteaya, that was "using same change-id" | 08:55 |
anteaya | well that was my take away when dhellmann encountered the same issue last week | 08:55 |
anteaya | from listening to jeblair help find a way forward | 08:55 |
AJaeger | and jeblair said "use same change-id only for backports", not for dependencies | 08:55 |
*** ddieterly has quit IRC | 08:56 | |
anteaya | how can you backport the same patch to both cinder and nova? | 08:56 |
*** Ala has joined #openstack-infra | 08:56 | |
anteaya | it isn't the same patch | 08:57 |
*** mrunge has quit IRC | 08:57 | |
markus_z | anteaya: In my case nova and cinder have both (98%) duplicated code regarding attaching/detaching volumes over fibre channel. | 08:58 |
anteaya | it isn't a backport | 08:59 |
anteaya | so shouldn't be using the same change id | 08:59 |
*** achuprin_ has joined #openstack-infra | 08:59 | |
anteaya | regardless | 08:59 |
anteaya | the depends-on won't work until the change id linked is unique | 08:59 |
*** mrunge has joined #openstack-infra | 09:00 | |
markus_z | right, I just wanted to construct a case where it would be possible. | 09:00 |
anteaya | where it would be possible | 09:00 |
anteaya | where what would be possible? | 09:00 |
markus_z | A backport of the same patch to two different projects. But that was just theoretically. | 09:00 |
anteaya | I don't know as my definition of backport is matching your definition of backport | 09:01 |
anteaya | and for depends-on to work the change id of the link needs to be unique | 09:01 |
anteaya | as that is the only information zuul has to select the parent patch | 09:02 |
*** luqas has quit IRC | 09:02 | |
*** devvesa has joined #openstack-infra | 09:02 | |
*** baoli has joined #openstack-infra | 09:04 | |
markus_z | anteaya: The backport definition would be interesting but I don't have to solve this right now. The uniqueness is my current problem I try to solve. Is an afterwards change of the "change-id" of the abandoned review legit? Or will I cause trouble with this? | 09:04 |
anteaya | markus_z: can you be around in about 6 hours? | 09:05 |
markus_z | anteaya: Yes I can | 09:05 |
anteaya | markus_z: I would like to hear again from jeblair how he would prefer to fix a situation like this | 09:05 |
anteaya | as I don't want to give you suggestions which are unsupported by others | 09:06 |
markus_z | Sure, just ping me. Thanks for your time! | 09:07 |
anteaya | markus_z: better if you just return in 6 hours | 09:07 |
anteaya | as I'm not sure where I will be | 09:07 |
markus_z | OK | 09:07 |
anteaya | and jeblair is probably the best one to sort this out | 09:07 |
anteaya | thanks | 09:07 |
*** teran_ has quit IRC | 09:08 | |
*** dims has joined #openstack-infra | 09:11 | |
*** dizquierdo has joined #openstack-infra | 09:11 | |
*** dims has quit IRC | 09:18 | |
*** AJaeger has quit IRC | 09:18 | |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add initial compute functional tests to Shade https://review.openstack.org/158463 | 09:22 |
*** markvoelker has quit IRC | 09:22 | |
*** krisgan has quit IRC | 09:25 | |
*** mfink_ has joined #openstack-infra | 09:25 | |
*** che-arne has joined #openstack-infra | 09:26 | |
*** mfink_ has quit IRC | 09:27 | |
*** mfink_ has joined #openstack-infra | 09:27 | |
*** zz_johnthetubagu is now known as johnthetubaguy | 09:27 | |
*** david-lyle_afk has quit IRC | 09:30 | |
*** david-lyle_afk has joined #openstack-infra | 09:30 | |
prmtl | folks, could someone point me to some docs describing process of renaming stackforge repo? | 09:30 |
*** AJaeger has joined #openstack-infra | 09:30 | |
prmtl | I saw some threads on mailing lists about renaming, is this the only source of info? | 09:31 |
openstackgerrit | Sumit Naiksatam proposed openstack-infra/project-config: Create experimental job for group-based-policy https://review.openstack.org/161511 | 09:31 |
*** Daisy__ has quit IRC | 09:32 | |
*** jistr has quit IRC | 09:34 | |
*** yamamoto has quit IRC | 09:35 | |
*** ujuc has quit IRC | 09:36 | |
*** ihrachyshka has joined #openstack-infra | 09:38 | |
prmtl | ok, I've found a doc: http://ci.openstack.org/gerrit.html#renaming-a-project | 09:39 |
prmtl | so what actions are required from me to init a rename process? what about stackalitics? | 09:40 |
*** teran has joined #openstack-infra | 09:42 | |
*** teran_ has joined #openstack-infra | 09:45 | |
*** jistr has joined #openstack-infra | 09:46 | |
*** otter768 has joined #openstack-infra | 09:46 | |
*** teran has quit IRC | 09:46 | |
*** amotoki_ has joined #openstack-infra | 09:48 | |
*** amotoki_ has quit IRC | 09:48 | |
*** vilobhmm has quit IRC | 09:48 | |
*** otter768 has quit IRC | 09:51 | |
*** ddieterly has joined #openstack-infra | 09:52 | |
*** afazekas has joined #openstack-infra | 09:52 | |
*** pelix has joined #openstack-infra | 09:56 | |
*** dims has joined #openstack-infra | 09:56 | |
*** ddieterly has quit IRC | 09:56 | |
*** mfink_ has quit IRC | 09:57 | |
*** ssam2 has joined #openstack-infra | 09:57 | |
openstackgerrit | Thomas Goirand proposed openstack/requirements: Typoe in LrDragNDrop https://review.openstack.org/162936 | 09:59 |
*** krisgan has joined #openstack-infra | 10:02 | |
*** mpaolino has quit IRC | 10:02 | |
*** jp_at_hp has joined #openstack-infra | 10:08 | |
*** Ala has quit IRC | 10:08 | |
*** sdake_ has quit IRC | 10:09 | |
*** faizan has joined #openstack-infra | 10:13 | |
*** sdake_ has joined #openstack-infra | 10:14 | |
openstackgerrit | yolanda.robla proposed openstack-infra/puppet-jenkins: Add configuration for HipChat in jjb https://review.openstack.org/162152 | 10:14 |
faizan | anteaya: Hi | 10:14 |
*** e0ne has joined #openstack-infra | 10:15 | |
faizan | anteaya: I need some help on this review for ironic-lib - https://review.openstack.org/#/c/161600/ | 10:15 |
*** mpaolino has joined #openstack-infra | 10:15 | |
*** chlong has joined #openstack-infra | 10:18 | |
*** ociuhandu has quit IRC | 10:20 | |
AJaeger | prmtl, see https://review.openstack.org/160963 as example | 10:21 |
AJaeger | prmtl, If you want to go into openstack namespace, you need a change for governance repository approved | 10:21 |
*** dhritishikhar has joined #openstack-infra | 10:22 | |
AJaeger | faizan, better check who's core reviewer in a project before you ping people. Go to your review, click on the project name - this will lead to https://review.openstack.org/#/admin/projects/openstack-infra/devstack-gate - click on Access - https://review.openstack.org/#/admin/projects/openstack-infra/devstack-gate,access and then on the core team for list of members | 10:23 |
*** ghostpl_ has joined #openstack-infra | 10:24 | |
faizan | AJaeger: Thanks for the pointers. | 10:26 |
openstackgerrit | Julien Danjou proposed openstack-infra/project-config: Move Gnocchi from Stackforge to OpenStack https://review.openstack.org/162146 | 10:31 |
*** coolsvap is now known as coolsvap|afk | 10:33 | |
*** yfried_ is now known as yfried|afk | 10:34 | |
*** yfried|afk is now known as yfried_ | 10:38 | |
*** bhuvan has quit IRC | 10:47 | |
*** salv-orlando has joined #openstack-infra | 10:48 | |
openstackgerrit | Merged openstack-infra/jenkins-job-builder: Add support to customize trigger-builds block thresholds https://review.openstack.org/154379 | 10:48 |
openstackgerrit | Sebastian Kalinowski proposed openstack-infra/project-config: Rename fuel-tasks-validator to fuel-tasklib https://review.openstack.org/162952 | 10:48 |
prmtl | AJaeger: thx | 10:48 |
*** hugh_afk has quit IRC | 10:50 | |
*** sdake_ has quit IRC | 10:51 | |
*** chlong has quit IRC | 10:51 | |
*** hughhalf has joined #openstack-infra | 10:51 | |
*** ddieterly has joined #openstack-infra | 10:53 | |
*** markvoelker has joined #openstack-infra | 10:53 | |
*** salv-orlando has quit IRC | 10:53 | |
*** baoli has quit IRC | 10:54 | |
*** vponomaryov1 has quit IRC | 10:54 | |
*** vponomaryov has joined #openstack-infra | 10:55 | |
*** salv-orlando has joined #openstack-infra | 10:55 | |
*** ddieterly has quit IRC | 10:58 | |
*** pc_m has joined #openstack-infra | 11:00 | |
*** mpaolino has quit IRC | 11:07 | |
*** alexpilotti_ has joined #openstack-infra | 11:08 | |
*** Somay has joined #openstack-infra | 11:09 | |
*** e0ne is now known as e0ne_ | 11:09 | |
*** e0ne_ is now known as e0ne | 11:09 | |
*** Krast has quit IRC | 11:09 | |
*** alexpilotti has quit IRC | 11:10 | |
*** alexpilotti_ is now known as alexpilotti | 11:10 | |
openstackgerrit | yolanda.robla proposed openstack-infra/puppet-logstash: Fixes for logstash manifests. https://review.openstack.org/162965 | 11:13 |
*** Longgeek has quit IRC | 11:14 | |
openstackgerrit | yolanda.robla proposed openstack-infra/system-config: Call logstash parent classes https://review.openstack.org/162966 | 11:20 |
openstackgerrit | yolanda.robla proposed openstack-infra/puppet-logstash: Fixes for logstash manifests. https://review.openstack.org/162965 | 11:20 |
openstackgerrit | Merged openstack-infra/jenkins-job-builder: Support additional features in Build Flow project https://review.openstack.org/158084 | 11:20 |
*** achuprin_ has quit IRC | 11:20 | |
*** mpaolino has joined #openstack-infra | 11:20 | |
*** alexpilotti has quit IRC | 11:21 | |
*** vponomaryov has quit IRC | 11:22 | |
*** vponomaryov has joined #openstack-infra | 11:22 | |
*** vponomaryov has quit IRC | 11:29 | |
*** jlanoux_ has joined #openstack-infra | 11:29 | |
*** vponomaryov has joined #openstack-infra | 11:29 | |
*** jlanoux has quit IRC | 11:32 | |
*** adalbas has joined #openstack-infra | 11:34 | |
openstackgerrit | Matthias Runge proposed openstack/requirements: Raise cap for Django https://review.openstack.org/155353 | 11:34 |
openstackgerrit | Victor Ryzhenkin proposed openstack-infra/project-config: Fix for gate-murano-congress-devstack-dsvm https://review.openstack.org/162971 | 11:35 |
BobBall | asselin, jeblair: Possible lead on NoneType from zuul - 162779,1 seems to trigger it for me. This change was created+abandoned very shortly after but also has a cross project dependency on a project that we do not have any rules for. | 11:36 |
*** Longgeek has joined #openstack-infra | 11:36 | |
*** ldnunes has joined #openstack-infra | 11:39 | |
*** mfink_ has joined #openstack-infra | 11:41 | |
*** _shaps_ has joined #openstack-infra | 11:41 | |
*** cdent has joined #openstack-infra | 11:43 | |
*** _shaps_ has quit IRC | 11:44 | |
*** mwagner_lap has quit IRC | 11:46 | |
*** jlanoux has joined #openstack-infra | 11:46 | |
*** mpaolino has quit IRC | 11:47 | |
*** otter768 has joined #openstack-infra | 11:47 | |
*** jlanoux_ has quit IRC | 11:49 | |
*** otter768 has quit IRC | 11:51 | |
*** _shaps_ has joined #openstack-infra | 11:51 | |
*** mmedvede has joined #openstack-infra | 11:52 | |
*** ddieterly has joined #openstack-infra | 11:54 | |
*** coolsvap|afk is now known as coolsvap | 11:54 | |
*** mrmartin has quit IRC | 11:56 | |
*** ddieterly has quit IRC | 11:58 | |
*** baoli has joined #openstack-infra | 11:58 | |
*** weshay has joined #openstack-infra | 12:01 | |
*** kgiusti has joined #openstack-infra | 12:01 | |
*** yfried_ is now known as yfried|afk | 12:02 | |
*** yfried|afk is now known as yfried_ | 12:05 | |
openstackgerrit | Aleksey Ripinen proposed openstack-infra/storyboard: Tasks branches, milestones and projects ids https://review.openstack.org/162126 | 12:11 |
*** dprince has joined #openstack-infra | 12:13 | |
*** BharatK has quit IRC | 12:14 | |
*** bswartz has quit IRC | 12:15 | |
*** sambetts has quit IRC | 12:16 | |
*** snikitin_ has quit IRC | 12:18 | |
*** johnthetubaguy is now known as zz_johnthetubagu | 12:19 | |
*** faizan has quit IRC | 12:19 | |
*** sambetts has joined #openstack-infra | 12:21 | |
*** dims has quit IRC | 12:25 | |
*** dims has joined #openstack-infra | 12:25 | |
*** mwagner_lap has joined #openstack-infra | 12:26 | |
*** aysyd has joined #openstack-infra | 12:27 | |
*** doug-fish has left #openstack-infra | 12:32 | |
*** mrmartin has joined #openstack-infra | 12:34 | |
*** markvoelker has quit IRC | 12:38 | |
*** markvoelker has joined #openstack-infra | 12:39 | |
*** gordc has joined #openstack-infra | 12:41 | |
*** markvoelker has quit IRC | 12:43 | |
*** mwagner_lap has quit IRC | 12:43 | |
*** rlandy has joined #openstack-infra | 12:44 | |
*** mwagner_lap has joined #openstack-infra | 12:44 | |
*** jistr has quit IRC | 12:47 | |
*** jistr has joined #openstack-infra | 12:48 | |
*** annegentle has joined #openstack-infra | 12:51 | |
*** ihrachyshka has quit IRC | 12:51 | |
*** ihrachyshka has joined #openstack-infra | 12:51 | |
*** yfried_ is now known as yfried|afk | 12:54 | |
*** ddieterly has joined #openstack-infra | 12:54 | |
*** bknudson has joined #openstack-infra | 12:56 | |
*** xyang1 has joined #openstack-infra | 12:56 | |
*** ddieterly has quit IRC | 12:56 | |
*** mpaolino has joined #openstack-infra | 12:58 | |
*** dboik has joined #openstack-infra | 12:59 | |
*** bhuvan has joined #openstack-infra | 12:59 | |
*** bhuvan has quit IRC | 13:04 | |
*** zul has quit IRC | 13:04 | |
*** jistr has quit IRC | 13:04 | |
*** emagana has joined #openstack-infra | 13:05 | |
*** jistr has joined #openstack-infra | 13:05 | |
*** coolsvap is now known as coolsvap|afk | 13:06 | |
*** stevemar has joined #openstack-infra | 13:06 | |
*** vivcheri has joined #openstack-infra | 13:09 | |
*** zul has joined #openstack-infra | 13:09 | |
*** banix has joined #openstack-infra | 13:12 | |
*** _shaps_ has quit IRC | 13:12 | |
*** amotoki has quit IRC | 13:12 | |
*** markvoelker has joined #openstack-infra | 13:13 | |
*** fifieldt has joined #openstack-infra | 13:14 | |
*** yfried|afk is now known as yfried_ | 13:14 | |
banix | AJaeger: Hi; I had to rebase the patch you had +2’ed. Please have a look again when you get a chance: https://review.openstack.org/#/c/162681/ Thank you. | 13:15 |
*** dkliban_afk is now known as dkliban | 13:16 | |
*** tnovacik has quit IRC | 13:17 | |
*** macjack has quit IRC | 13:17 | |
AJaeger | banix, done | 13:18 |
banix | AJaeger: thank you. | 13:18 |
*** stevemar has quit IRC | 13:18 | |
*** stevemar has joined #openstack-infra | 13:19 | |
*** Somay has quit IRC | 13:19 | |
*** peristeri has joined #openstack-infra | 13:20 | |
*** dustins has joined #openstack-infra | 13:22 | |
*** doug-fish has joined #openstack-infra | 13:24 | |
*** ildikov has quit IRC | 13:24 | |
*** yfried_ is now known as yfried|afk | 13:24 | |
openstackgerrit | Matthew Treinish proposed openstack-infra/project-config: Add skip if for tempest running tempest on tempest commits https://review.openstack.org/162993 | 13:24 |
mtreinish | jogo: ^^^ I took your advice on doing that | 13:25 |
*** sigmavirus24_awa is now known as sigmavirus24 | 13:25 | |
*** changbl has quit IRC | 13:26 | |
*** rkukura has quit IRC | 13:27 | |
mtreinish | it also helps that I had a bunch of +Ad docs patches I have to recheck because pep8 was blocked yesterday | 13:28 |
*** yamamoto has joined #openstack-infra | 13:28 | |
sigmavirus24 | anteaya: if you still want to see the glance patches, let me know. I can dig them up (one was mine) | 13:28 |
*** yamamoto has quit IRC | 13:29 | |
*** yamamoto has joined #openstack-infra | 13:29 | |
mtreinish | jeblair, fungi: do you guys have a script which runs recheck on everything in a project? | 13:30 |
*** Ala has joined #openstack-infra | 13:31 | |
*** amitgandhinz has joined #openstack-infra | 13:31 | |
fungi | mtreinish: not really, no | 13:32 |
fungi | mtreinish: zuul has a command-line interface which can reenqueue arbitrary changes | 13:32 |
fungi | mtreinish: but we still have to do something to build the list of changes to feed into it | 13:32 |
mtreinish | fungi: hmm, ok. I guess I'll just ML post to tell people to recheck if pep8 failed yesterday | 13:32 |
fungi | if it's a problem in one project, i usually do a quick gerrit query for changes in that project with a -1 or -2 verify from jenkins and manually recheck if they failed the one job which i'm aware was broken | 13:33 |
fungi | assuming the number impacted is erasonably small (say <25) | 13:34 |
*** ildikov has joined #openstack-infra | 13:34 | |
fungi | reasonably too | 13:34 |
*** yfried|afk is now known as yfried_ | 13:34 | |
*** Ala has quit IRC | 13:34 | |
annegentle | mtreinish: what happened with doc8 yesterday? | 13:35 |
mtreinish | fungi: yeah it should be a pretty small set of patches, I guess it'll be easy to query it off | 13:35 |
*** mattfarina has joined #openstack-infra | 13:35 | |
*** tiswanso has joined #openstack-infra | 13:36 | |
*** craigbr has joined #openstack-infra | 13:36 | |
mtreinish | annegentle: was something up with doc8 yesterday? We had the tempest queue die yesterday because of the sample config generator in the pep8 job breaking with the oslo.config release yesterday | 13:36 |
*** rkukura has joined #openstack-infra | 13:37 | |
fungi | mrmartin: oh, i built the replacement ask server yesterday. it's up at 104.239.140.123/2001:4800:7819:104:be76:4eff:fe05:b1be but i haven't tried it out or performed any of the post-install steps yet to import the old data | 13:39 |
mrmartin | fungi, thanks, I'll review that. | 13:40 |
*** Ala has joined #openstack-infra | 13:40 | |
*** garyh has joined #openstack-infra | 13:40 | |
banix | jeblair: Hi; I had to rebase the patch you had +2’ed. Please have a look again when/if you get a chance: https://review.openstack.org/#/c/162681/ Thank you. | 13:42 |
annegentle | mtreinish: ah, okay. | 13:43 |
*** ociuhandu has joined #openstack-infra | 13:45 | |
*** salv-orlando has quit IRC | 13:46 | |
*** armax has joined #openstack-infra | 13:46 | |
*** ddieterly has joined #openstack-infra | 13:46 | |
*** mriedem has joined #openstack-infra | 13:48 | |
*** otter768 has joined #openstack-infra | 13:48 | |
*** salv-orlando has joined #openstack-infra | 13:48 | |
openstackgerrit | Sudheesh proposed openstack-infra/project-config: openstack docker - this is a coobook that will configure docker capabalities into openstack services https://review.openstack.org/162528 | 13:48 |
*** sdake_ has joined #openstack-infra | 13:49 | |
*** annegentle has quit IRC | 13:51 | |
*** ayoung has joined #openstack-infra | 13:51 | |
*** mriedem has quit IRC | 13:52 | |
*** craigbr has quit IRC | 13:52 | |
*** otter768 has quit IRC | 13:53 | |
*** peristeri has quit IRC | 13:53 | |
*** mtanino has joined #openstack-infra | 13:54 | |
*** mriedem has joined #openstack-infra | 13:56 | |
*** yamahata has quit IRC | 13:56 | |
*** yamahata has joined #openstack-infra | 13:57 | |
*** dkranz has joined #openstack-infra | 13:57 | |
*** r-daneel has joined #openstack-infra | 13:58 | |
*** bswartz has joined #openstack-infra | 13:58 | |
*** marun has joined #openstack-infra | 13:58 | |
*** carl_baldwin has joined #openstack-infra | 13:59 | |
*** craigbr has joined #openstack-infra | 13:59 | |
*** ominakov_ has joined #openstack-infra | 14:00 | |
*** hdd_ has joined #openstack-infra | 14:01 | |
*** ayoung has quit IRC | 14:02 | |
*** ominakov__ has joined #openstack-infra | 14:02 | |
*** ominakov_ has quit IRC | 14:04 | |
*** dhritishikhar has quit IRC | 14:05 | |
openstackgerrit | yolanda.robla proposed openstack-infra/system-config: Add a cronjob frequency to pypi_mirror https://review.openstack.org/163014 | 14:06 |
*** _nadya_ has joined #openstack-infra | 14:07 | |
*** annegentle has joined #openstack-infra | 14:08 | |
anteaya | sigmavirus24: thanks I think we figured out that open connections retain credentials through the gerrit cache, we will clear gerrit caches next time we have to disable | 14:09 |
anteaya | sigmavirus24: thanks for following up, I appreciate it | 14:09 |
*** tnovacik has joined #openstack-infra | 14:10 | |
sigmavirus24 | No worries. | 14:11 |
*** annegentle has quit IRC | 14:11 | |
*** ildikov has quit IRC | 14:11 | |
fungi | mrmartin: also, i built the new server with 4gb ram rather than 2gb, to allow for additional growth, but we may need to think about disk space. new instances are coming with 40g on / and 40g on /opt rather than 80g on / (but for now the current utilization isn't at risk for overrunning / anyway) | 14:13 |
*** wenlock has joined #openstack-infra | 14:13 | |
*** craigbr has quit IRC | 14:14 | |
*** yfried__ has joined #openstack-infra | 14:14 | |
*** yfried_ has quit IRC | 14:15 | |
dougwig | good morning (or evening) infra. can i get a second core to look at this review to add an api/tempest experimental job for lbaas? https://review.openstack.org/#/c/161432/ | 14:15 |
*** timcline has joined #openstack-infra | 14:16 | |
*** timcline has quit IRC | 14:16 | |
*** garyk has quit IRC | 14:16 | |
*** timcline has joined #openstack-infra | 14:17 | |
*** BharatK has joined #openstack-infra | 14:18 | |
*** cdent has quit IRC | 14:18 | |
*** cdent has joined #openstack-infra | 14:18 | |
*** garyk has joined #openstack-infra | 14:20 | |
*** eharney has joined #openstack-infra | 14:21 | |
*** prmtl has quit IRC | 14:21 | |
*** amotoki has joined #openstack-infra | 14:21 | |
*** prad has joined #openstack-infra | 14:23 | |
*** zz_johnthetubagu is now known as johnthetubaguy | 14:23 | |
*** ChuckC has quit IRC | 14:23 | |
openstackgerrit | Merged openstack-infra/project-config: New job for in-tree lbaasv2 tempest tests https://review.openstack.org/161432 | 14:26 |
*** ildikov has joined #openstack-infra | 14:27 | |
*** mpaolino has quit IRC | 14:27 | |
fungi | dougwig: ^ | 14:29 |
dougwig | fungi: thank you! | 14:30 |
*** Guest75125 has joined #openstack-infra | 14:30 | |
*** jlk has joined #openstack-infra | 14:31 | |
*** emagana has quit IRC | 14:31 | |
*** ominakov__ has quit IRC | 14:31 | |
jlk | mordred: hi there. I'm pointed in your direction for some help with packaging, and pbr, and things like nova-manage being able to find the version at runtime of things. | 14:31 |
jlk | maybe you could give a hand. We've got python venvs wrapped into apt packages. Service (nova) seems to run fine, but nova-manage crashes out when it tries to import nova, which tries to import glanceclient which tries to figure out __version__ which calls into pbr, which tries to open a relative path to PKG-INFO which fails, and then I get a traceback. | 14:32 |
*** achanda has joined #openstack-infra | 14:34 | |
anteaya | BobBall: can you join #openstack-meeting? neutron is discussing xen and resources | 14:34 |
*** zz_jgrimm is now known as jgrimm | 14:35 | |
*** carl_baldwin_ has joined #openstack-infra | 14:35 | |
*** carl_baldwin has quit IRC | 14:37 | |
*** carl_baldwin_ is now known as carl_baldwin | 14:37 | |
*** markvoelker has quit IRC | 14:37 | |
fungi | jlk: how did you install things into the virtualenv? via pip or something else? | 14:37 |
jlk | fungi: pip I believe. | 14:37 |
*** Somay has joined #openstack-infra | 14:37 | |
*** markvoelker has joined #openstack-infra | 14:37 | |
openstackgerrit | Sudheesh proposed openstack-infra/project-config: openstack docker - single ammended final review request https://review.openstack.org/162528 | 14:38 |
*** bhunter71 has joined #openstack-infra | 14:38 | |
fungi | jlk: have a traceback you can post to paste.o.o? | 14:39 |
*** garyk has quit IRC | 14:39 | |
*** pblaho has joined #openstack-infra | 14:40 | |
mordred | jlk: ola! | 14:40 |
*** dhritishikhar has joined #openstack-infra | 14:41 | |
jlk | mordred: https://gist.github.com/j2sol/b606c5eecef385791d00 | 14:41 |
*** markvoelker has quit IRC | 14:42 | |
*** hdd_ has quit IRC | 14:42 | |
mordred | jlk: it sounds like a pkg-resources issue with the relocated venv ... Looking | 14:42 |
*** achanda has quit IRC | 14:43 | |
dougwig | can someone point me at the source for the requirements proposal bot? a google search for the likely terms just finds me jabillions of its submitted reviews. | 14:43 |
*** woodster_ has joined #openstack-infra | 14:43 | |
fungi | ahh, yeah moving virtualenvs around results in a lot of broken without some careful path mangling | 14:43 |
*** Guest75125 is now known as notsogentle | 14:43 | |
*** notsogentle is now known as agentle_ | 14:43 | |
openstackgerrit | Doug Hellmann proposed openstack/requirements: Cap Oslo libraries for kilo https://review.openstack.org/162656 | 14:43 |
fungi | requiring directly editing references embedded within the virtualenv | 14:43 |
*** rwsu has joined #openstack-infra | 14:44 | |
jeblair | markus_z, anteaya: it is certainly fine to use the same change id across projects if it represents the same logical change (eg -- updating tox.ini everywhere). if they are different changes, don't use the same change-id just for depends-on -- you can add two depends-on lines. | 14:44 |
BobBall | thanks for the heads up anteaya | 14:44 |
*** dangers_away is now known as dangers | 14:44 | |
anteaya | BobBall: of course | 14:45 |
markus_z | jeblair: Thanks, I will take care of that in future patch sets. | 14:45 |
fungi | dougwig: it's not really a bot, just a jenkins job running a script at http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/scripts/propose_update.sh | 14:45 |
anteaya | BobBall: you do good work, I'm doing my best to support that | 14:45 |
jeblair | markus_z, anteaya: currently if a change depends-on an abandoned change, there is no way to get it to progress in zuul. i think the preferred option right now would be to simply remove the depends-on from the commit and make sure things merge in the correct order manually. | 14:45 |
dougwig | fungi: oh, that's right. thank you. | 14:45 |
anteaya | jeblair: thank you | 14:45 |
jeblair | markus_z, anteaya: or, you could upload the changes again with a new change-id. | 14:45 |
BobBall | :) Appreciated. | 14:45 |
*** mfink_ has quit IRC | 14:45 | |
jeblair | markus_z, anteaya: i expect we'll change something in zuul's behavior to deal with this better soon. | 14:45 |
mordred | jlk: I'm on my phone ... May take me a minute to figure out ... Is there a way I can 'easily' reproduce? | 14:46 |
*** nelsnelson has joined #openstack-infra | 14:46 | |
anteaya | jeblair: thanks for including me in the info, I appreciate it | 14:46 |
anteaya | BobBall: :) | 14:46 |
jlk | mordred: take all the time you need, I'm just sitting next to you. | 14:46 |
*** zz_dimtruck is now known as dimtruck | 14:46 | |
fungi | heh | 14:46 |
*** yolanda has quit IRC | 14:48 | |
*** markvoelker has joined #openstack-infra | 14:49 | |
jeblair | dhellmann: why did you abandon https://review.openstack.org/#/c/95885/6 ? | 14:49 |
*** yolanda has joined #openstack-infra | 14:50 | |
markus_z | jeblair: Is there a way I can remove a change-id from an abandoned patch? | 14:51 |
jeblair | (also i did not know it was abandoned -- i think we need to establish some kind of process around abandoning things like that so that people are informed) | 14:51 |
markus_z | To ensure the uniqueness of my other, real patch. | 14:51 |
fungi | looks like the abandonment was based on an assumption that infra follows the integrated release cycle, and also the rather annoying pattern in other projects of requiring people to repropose unapproved specs each cycle | 14:51 |
fungi | which strikes me as a lot of busywork | 14:52 |
jeblair | markus_z: no, it's permanent, you would have to re-upload the changes you want with new change-ids | 14:52 |
jeblair | fungi: yes, it's certainly not the case here | 14:52 |
markus_z | jeblair: OK, then in my case I assume it makes sense to remove the "depends-on" and watch over it manually. Thanks for clarifying. | 14:53 |
fungi | markus_z: a change in gerrit is _defined_ by its change-id+project+branch so every change must have a change-id and altering the change-id means it's a new change | 14:53 |
*** jaypipes has quit IRC | 14:53 | |
*** e0ne is now known as e0ne_ | 14:54 | |
*** eharney has quit IRC | 14:54 | |
markus_z | fungi: That was my assumption as well, thank you. I was hoping for some black-magic trick I didn't know yet :) | 14:54 |
jeblair | fungi, dhellmann: fyi, my thought process was: why are oslo releases still breaking things when we have a test infrastructure that can prevent this? let me go look at the oslo test spec and see what part of it is broken. wait, where is the oslo test spec? i'm sure i +2d it.... | 14:55 |
clarkb | good morning. I restarted etherpad-dev's etherpad process last night to get it running on the latest development version please test. I have a specific etherpad started if you want to poke around but general use would be good https://etherpad-dev.openstack.org/p/clarkb-testing | 14:55 |
openstackgerrit | Sergey Lukjanov proposed openstack/requirements: Bump sahara client version https://review.openstack.org/155428 | 14:55 |
clarkb | also jenkins02 had ~4k threads last night so I expect it to be the next to go. | 14:56 |
*** hashar has quit IRC | 14:56 | |
*** hashar has joined #openstack-infra | 14:57 | |
dhellmann | jeblair: it wasn't going to land (I wasn't going to finish the work) and at some point sdague convinced me we wouldn't have the resources to run that many jobs (think about how many projects rely on oslo.config) | 14:57 |
*** e0ne has joined #openstack-infra | 14:57 | |
anteaya | clarkb: thanks, I am testing | 14:57 |
anteaya | not sure what else to do to test an ether pad | 14:57 |
clarkb | anteaya: I think mostly connect to it and do things that you would normally do. I need to confirm chat still works for example | 14:58 |
*** eharney has joined #openstack-infra | 14:58 | |
*** dims has quit IRC | 14:58 | |
*** carl_baldwin has quit IRC | 14:59 | |
anteaya | okay connected and working on remember what I normally do | 14:59 |
anteaya | remembering | 14:59 |
jeblair | dhellmann: next time can you let the rest of us know? :) | 15:00 |
*** stevemar2 has joined #openstack-infra | 15:01 | |
*** stevemar has quit IRC | 15:01 | |
*** dimsum__ has joined #openstack-infra | 15:01 | |
dhellmann | jeblair: what's the best way to do that? I thought abandoning the review would be a clear sign? | 15:02 |
*** achanda has joined #openstack-infra | 15:03 | |
jeblair | dhellmann: yes, but i did not see the sign until 7 months after it was abandoned... i honestly don't know, but we should figure something out.... maybe a mail to the list, or an item at the meeting? perhaps as ptl, i should just start searching for abandoned specs every few weeks and announce them? | 15:04 |
*** ChuckC has joined #openstack-infra | 15:04 | |
jeblair | dhellmann: i'm not trying to pick on you here... this is a general problem that needs solving | 15:04 |
anteaya | clarkb: the test etherpad works as expected for me | 15:04 |
dhellmann | jeblair: I guess I assumed you would see the notice from it being abandoned, but you probably get even more gerrit email than I do | 15:04 |
dhellmann | jeblair: yeah, definitely | 15:05 |
*** agentle_ has quit IRC | 15:05 | |
jeblair | dhellmann: i have 14,944 unread emails from gerrit | 15:05 |
dhellmann | jeblair: untennable | 15:05 |
clarkb | anteaya: thank you for checking | 15:05 |
*** pblaho_ has joined #openstack-infra | 15:05 | |
*** achanda has quit IRC | 15:05 | |
anteaya | sure, thanks for setting it up | 15:05 |
*** pblaho has quit IRC | 15:06 | |
clarkb | its likely we will want to put that specific version on the etherpad.o.o server as soon as ops meetup is over | 15:06 |
*** mpavone_ has quit IRC | 15:06 | |
jeblair | dhellmann: it only goes up :) | 15:07 |
jeblair | dhellmann: let me see if i can alter my email preferences to something useful | 15:07 |
dhellmann | jeblair: yeah, I have fastmail set to auto-delete any that are older than 60 days | 15:07 |
*** vilobhmm has joined #openstack-infra | 15:08 | |
jeblair | dhellmann: though i stopped 'watching' projects after i wrote gertty, so i'm actually not getting email for about 70 projects which i review.... | 15:08 |
dhellmann | jeblair: so, do you really think there's a workable way to set up that cross-project unit testing? I'm not as enthusiastic about it as I was. The most recent set of releases included ~12 libs and we had 2 things break, one was a legit bug and one was a side-effect of incubation. | 15:08 |
dhellmann | both only affected one project each, afaict | 15:08 |
dhellmann | jeblair: I need to spend more time with gerrty | 15:09 |
jeblair | dhellmann: things are certainly different since the last time we discussed it. i don't have an answer right now, but i will think about it. | 15:09 |
*** hyakuhei has joined #openstack-infra | 15:10 | |
hyakuhei | Hi Infra, anyone around? | 15:10 |
*** ildikov has quit IRC | 15:10 | |
dhellmann | jeblair: ok | 15:10 |
jeblair | dhellmann: omg -- i got >250 emails from gerrit in the last 24 hours, and i have email notifications set up for _zero_ projects. | 15:11 |
dhellmann | jeblair: yeah, I wish there was a way to turn it all off entirely | 15:12 |
jeblair | i think that's _just_ comments on changes where i'm a reviewer | 15:12 |
* dhellmann wonders if fastmail thinks our gerrit server is a spam bot | 15:12 | |
*** katyafervent has quit IRC | 15:12 | |
*** dteselkin has quit IRC | 15:12 | |
*** rakhmerov has quit IRC | 15:12 | |
*** tsufiev has quit IRC | 15:12 | |
*** vilobhmm has left #openstack-infra | 15:12 | |
*** tsufiev has joined #openstack-infra | 15:13 | |
*** ominakov_ has joined #openstack-infra | 15:13 | |
*** garyk has joined #openstack-infra | 15:13 | |
*** tsufiev has quit IRC | 15:14 | |
*** yfried has joined #openstack-infra | 15:14 | |
dhellmann | jeblair: this script might be of interest (we were using it by hand for the namespace package change pre-release testing) http://git.openstack.org/cgit/openstack/oslotest/tree/tools/oslo_run_pre_release_tests | 15:15 |
anteaya | hyakuhei: this channel is logged so it is best to just ask your question since even if noone happens to be around at the time you ask your question someone often may respond later | 15:15 |
*** radez_g0n3 is now known as radez | 15:15 | |
mordred | fungi: just fyi - the jlk issue was old pip/setuptools | 15:15 |
anteaya | hyakuhei: you can find the channel logs here: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/ | 15:15 |
*** dkranz has quit IRC | 15:16 | |
fungi | mordred: sounds likely... and site packages enabled in the venv presumably? | 15:16 |
*** unicell has joined #openstack-infra | 15:17 | |
*** yfried__ has quit IRC | 15:17 | |
mordred | fungi: (we looked at it - sitting next to each other) | 15:17 |
mordred | fungi: it was that python-glanceclient was installed from a wheel, but the setuptools in the venv didn't know how to process the pkg_resources dist-info file that wheels install | 15:17 |
*** yfried_ has joined #openstack-infra | 15:17 | |
*** unicell1 has quit IRC | 15:18 | |
*** yfried has quit IRC | 15:18 | |
fungi | awesome | 15:18 |
*** phschwartz has quit IRC | 15:18 | |
*** phschwartz has joined #openstack-infra | 15:19 | |
*** rakhmerov has joined #openstack-infra | 15:20 | |
*** hyakuhei has quit IRC | 15:20 | |
*** dannywilson has joined #openstack-infra | 15:20 | |
*** Somay has quit IRC | 15:21 | |
*** dannywilson has quit IRC | 15:21 | |
*** dteselkin has joined #openstack-infra | 15:21 | |
*** dannywilson has joined #openstack-infra | 15:21 | |
*** boris-42 has quit IRC | 15:22 | |
*** jcoufal has quit IRC | 15:22 | |
*** david-lyle_afk is now known as david-lyle | 15:23 | |
*** che-arne has quit IRC | 15:23 | |
*** katyafervent has joined #openstack-infra | 15:23 | |
clarkb | https://review.openstack.org/#/c/159159/ https://review.openstack.org/#/c/154305/ and https://review.openstack.org/#/c/154300/ should together form the fix for the ansible puppet new project race | 15:24 |
*** tsufiev_ has joined #openstack-infra | 15:24 | |
*** tnovacik has quit IRC | 15:25 | |
clarkb | I got the base run ansible like ansible changes mordred wrote merged last week which allows us to move forward on using ansible-puppet | 15:25 |
mordred | fungi: you'll love it | 15:25 |
*** imcsk8 has quit IRC | 15:26 | |
*** imcsk8 has joined #openstack-infra | 15:26 | |
*** Somay has joined #openstack-infra | 15:26 | |
fungi | mordred: i'm loving it already and i haven't even looked at it yet ;) | 15:27 |
*** purp_away is now known as purp | 15:27 | |
clarkb | mordred: do I need quotes around the {{}}s on lines 17 and 24 https://review.openstack.org/#/c/154305/4/modules/openstack_project/files/ansible/playbooks/remote_puppet_git.yaml ? | 15:28 |
fungi | clarkb: i'm going to test our timing hypothesis for the puppet uid/gid race by standing up a dummy gerrit server and then poking around in it. that's the manifest where i originally observed the problem which led to the login.defs change | 15:28 |
*** emagana has joined #openstack-infra | 15:28 | |
clarkb | fungi: sounds good | 15:28 |
jlk | clarkb: I think you would. { after a : is generally a thing that makes yaml upset. | 15:29 |
clarkb | jlk: it depends on when ansible applies applies the templating | 15:29 |
fungi | not necessarily upset, it just assumes you're defining a literal map | 15:29 |
clarkb | jlk: if before parsed by yaml it should be fine, if after it may cause problems | 15:29 |
*** ChuckC has quit IRC | 15:30 | |
clarkb | but I can add them as quotes should be fine in both cases | 15:30 |
*** ujuc has joined #openstack-infra | 15:30 | |
*** sdake__ has joined #openstack-infra | 15:30 | |
*** che-arne has joined #openstack-infra | 15:31 | |
*** purp is now known as purp_away | 15:31 | |
openstackgerrit | Merged openstack-infra/system-config: Document new puppet lockfile arg https://review.openstack.org/161889 | 15:31 |
mordred | clarkb: yes | 15:31 |
mordred | clarkb: you need the quotes | 15:32 |
mordred | clarkb: however, in a fit of amazing | 15:32 |
*** ominakov_ has quit IRC | 15:32 | |
mordred | clarkb: ansile will actualy tell you that in the error message | 15:32 |
*** carl_baldwin has joined #openstack-infra | 15:32 | |
mordred | clarkb: if you run it - it'll say "hey, this doesn't parse as yaml - it's probably because you need quotes" | 15:32 |
openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Pass in a ref for the project_config repo https://review.openstack.org/154305 | 15:33 |
mordred | clarkb: the templating is run on each value after the yaml has been parse, fwiw | 15:33 |
clarkb | mordred: ok so its like JJB | 15:33 |
mordred | yah | 15:33 |
*** carl_baldwin has quit IRC | 15:33 | |
*** sdake_ has quit IRC | 15:34 | |
clarkb | also the git module is a role or a task? | 15:35 |
*** changbl has joined #openstack-infra | 15:35 | |
clarkb | looks like it is a task? | 15:36 |
* clarkb figures out how to run some ansible out of a local venv | 15:36 | |
*** ujuc has quit IRC | 15:39 | |
*** sergsh has quit IRC | 15:39 | |
*** boris-42 has joined #openstack-infra | 15:39 | |
jlk | there is a git module | 15:40 |
jlk | you /may/ have a git role, but that'd be specific to whatever. | 15:40 |
jlk | (modules == tasks) | 15:40 |
clarkb | jlk: yup /me is figuring this out | 15:42 |
clarkb | also you MUST have an inventor file even with --connection=local :/ | 15:42 |
*** Ryan_Lane has joined #openstack-infra | 15:42 | |
*** esker has joined #openstack-infra | 15:42 | |
jlk | that I believe is changing in v2 | 15:42 |
jlk | hint | 15:42 |
jlk | ansible-playbook -i derp, | 15:42 |
jlk | you can specify raw inventory json to the -i option, by having a trailing comma. | 15:43 |
jlk | it's dirty, but quick. | 15:43 |
clarkb | wow | 15:43 |
clarkb | I got around it by making a file with only localhost in it and passing that path via -i | 15:43 |
jlk | that works as well. | 15:43 |
clarkb | I have mostly confirmed that the above change works (I replaced the puppet role with some echo gitinfo.after >> logfile | 15:44 |
clarkb | but ansible-playbook -v printed the after value too in any case I am happy with that change as is | 15:44 |
*** cody-somerville has joined #openstack-infra | 15:44 | |
*** gyee has joined #openstack-infra | 15:46 | |
clarkb | I am going to start jenkins02 surgery now, it appears to be mostly non responsive | 15:46 |
*** thedodd has joined #openstack-infra | 15:46 | |
jeblair | clarkb: thx | 15:46 |
*** changbl has quit IRC | 15:47 | |
*** yamamoto has quit IRC | 15:49 | |
*** otter768 has joined #openstack-infra | 15:49 | |
*** hdd_ has joined #openstack-infra | 15:49 | |
*** sergsh has joined #openstack-infra | 15:52 | |
*** ayoung has joined #openstack-infra | 15:52 | |
*** unicell has quit IRC | 15:53 | |
*** yamamoto has joined #openstack-infra | 15:53 | |
*** craigbr has joined #openstack-infra | 15:53 | |
*** unicell has joined #openstack-infra | 15:53 | |
clarkb | 02 is in shutdown mode with new plugin installed. Will restart when jobs are finished | 15:54 |
*** otter768 has quit IRC | 15:54 | |
*** sdake__ has quit IRC | 15:54 | |
clarkb | and since we are semi tracking it, jenkins02 has 4.5k threads | 15:54 |
*** sdake_ has joined #openstack-infra | 15:55 | |
clarkb | jenkins04 for comparison has 289 | 15:55 |
*** tonytan4ever has joined #openstack-infra | 15:56 | |
pleia2 | StevenK: how frustrating! and 1) I only suggested fedora/centos if it was easier, if you're going down a rabbit hole I think we'd rather it be Ubuntu 2) this is definitely something we can reach out to the Zanata guys at Red Hat for, they may even have someone assigned to rhel/centos packaging already | 15:56 |
*** banix has quit IRC | 15:56 | |
*** fifieldt has quit IRC | 15:57 | |
fungi | an interesting checklist for when people come looking for free software alternatives to proprietary/hosted saas stuff https://wiki.debian.org/FreedomBox/LeavingTheCloud | 15:57 |
anteaya | fungi: nice | 15:58 |
*** markvoelker has quit IRC | 15:59 | |
fungi | the file sharing section there lists a few i don't know we've looked into yet | 15:59 |
*** markvoelker has joined #openstack-infra | 16:00 | |
zaro | morning | 16:01 |
clarkb | zaro: good morning | 16:01 |
anteaya | zaro: morning | 16:01 |
zaro | clarkb: did you want to do anything about gerrit utf8 issue? | 16:01 |
anteaya | fungi: yes I only tried owncloud and seafiles | 16:01 |
*** Ryan_Lane has quit IRC | 16:01 | |
*** stevemar2 is now known as stevemar | 16:01 | |
clarkb | zaro: ya we should probably debug why those characters don't render properly | 16:01 |
*** koolhead17 has quit IRC | 16:01 | |
clarkb | zaro: I don't know that its very urgent but may indicate a bug somewhere | 16:02 |
asselin | BobBall, jeblair interesting observation. I remember observing something similar in a previous case, where the patch had e.g. a subsequent patch submitted before the first job finished (with cross project dependency and no rule) . We should submit a bug and track the details/occurences there. | 16:02 |
anteaya | and the one reed suggested a few weeks back, which was a locally running daemon | 16:02 |
*** banix has joined #openstack-infra | 16:02 | |
*** mrmartin has quit IRC | 16:02 | |
*** timcline has quit IRC | 16:02 | |
clarkb | zaro: also can you see my comment on 160549? | 16:02 |
fungi | clarkb: zaro: when i looked at the example comments, it appeared as if the non-ascii codepoints were being substituted, so not a typical mojibake situation and might be actual gerrit design/configuration at work? | 16:03 |
*** dkranz has joined #openstack-infra | 16:03 | |
*** changbl has joined #openstack-infra | 16:03 | |
zaro | clarkb: replied and i +1 that. | 16:03 |
fungi | or else i was looking at something different than everyone else was. it was sort of vague | 16:03 |
jeblair | i assume we're talking about "edit typo, '????' > '????'" | 16:03 |
*** yfried_ has quit IRC | 16:03 | |
clarkb | zaro: awesome thanks | 16:03 |
jeblair | which literally renders as question marks for me | 16:03 |
*** emagana has quit IRC | 16:04 | |
jeblair | i will look up the db values for that | 16:04 |
fungi | yeah, looks like what you get when you tell python to encode with substitution | 16:04 |
zaro | fungi: link? | 16:04 |
*** jcoufal has joined #openstack-infra | 16:04 | |
*** markvoelker has quit IRC | 16:05 | |
*** unicell has quit IRC | 16:05 | |
fungi | zaro: https://review.openstack.org/162045 (from that e-mail thread) | 16:05 |
fungi | obviously not python doing it in this case, but presumably java has similar encoding mechanisms | 16:06 |
pleia2 | I'm off to elasticon, cinerama will be giving the zanata update during the meeting later :) | 16:06 |
clarkb | pleia2: have fun | 16:07 |
*** tkelsey has quit IRC | 16:07 | |
clarkb | I did check the db default encoding last night is utf8 but I now realize I didn't check individual tables | 16:07 |
clarkb | maybe the comments table(s) avioded the default somehow? | 16:07 |
openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Add tests for yaml anchor behaviour https://review.openstack.org/158350 | 16:08 |
jeblair | in gsql that comment renders with question marks. will try again with mysql client directly | 16:08 |
nibalizer | good mornyng yall | 16:08 |
anteaya | morning nibalizer | 16:08 |
jeblair | ) ENGINE=InnoDB DEFAULT CHARSET=latin1 | 16:08 |
*** yamamoto has quit IRC | 16:08 | |
*** emagana_ has joined #openstack-infra | 16:08 | |
anteaya | that would do it | 16:08 |
*** ashleighfarnham has joined #openstack-infra | 16:09 | |
jeblair | i do not think we have all of the information | 16:09 |
anteaya | how have other languages been rendered? | 16:09 |
*** ashleighfarnham has quit IRC | 16:09 | |
*** ashleighfarnham has joined #openstack-infra | 16:09 | |
anteaya | surely korean can't be our first foray into this family of languages | 16:09 |
fungi | yup | 16:09 |
jeblair | keep in mind the charset of the table does not restrict what can be stored in it. | 16:10 |
*** agentle_ has joined #openstack-infra | 16:10 | |
jeblair | still literal ?'s in mysql client | 16:10 |
clarkb | jeblair: right it only affects ordering and globbing iirc | 16:11 |
mordred | Correct | 16:11 |
mordred | Also, storage size allocation | 16:11 |
zaro | here is this history with last discussion on converting to utf8: https://bugs.launchpad.net/openstack-ci/+bug/979227 | 16:11 |
openstack | Launchpad bug 979227 in OpenStack Core Infrastructure "convert gerrit's mysql tables to utf8" [Medium,Fix released] - Assigned to Khai Do (zaro0508) | 16:11 |
fungi | i wonder if gerrit encodes with substitution when storing into the db, perhaps inspecting the table's default charset to make that decision | 16:11 |
clarkb | but gerrit may try to accomodate the table encoding? | 16:11 |
clarkb | fungi: jinx | 16:11 |
fungi | yeah, that | 16:11 |
jeblair | https://review.openstack.org/#/c/156793/ | 16:12 |
jeblair | i copy/pasted some korean characters in there | 16:12 |
*** esker has quit IRC | 16:12 | |
jeblair | and a latin1 | 16:13 |
*** mrmartin has joined #openstack-infra | 16:13 | |
*** changbl has quit IRC | 16:13 | |
fungi | i followed up with some utf8 kanji and predictably get the same as for utf8 hangul | 16:15 |
openstackgerrit | Teemu Patja proposed stackforge/python-jenkins: Add a get_nodes method. https://review.openstack.org/163068 | 16:15 |
*** mpavone_ has joined #openstack-infra | 16:15 | |
*** vivcheri has quit IRC | 16:16 | |
*** arxcruz has quit IRC | 16:16 | |
*** emagana_ has quit IRC | 16:16 | |
*** ghostpl_ has quit IRC | 16:17 | |
*** emagana has joined #openstack-infra | 16:17 | |
*** sergsh has quit IRC | 16:17 | |
jeblair | https://code.google.com/p/gerrit/issues/detail?id=1101 | 16:18 |
jeblair | that lends support to the idea that the gerrit (via the db driver) is inspecting the table charset and modifying input to correspond | 16:18 |
*** ghostpl_ has joined #openstack-infra | 16:19 | |
anteaya | ihrachyshka: so when you are about to merge a patch that will break ci systems, perhaps the next time you could consult with us? | 16:20 |
*** banix has quit IRC | 16:21 | |
anteaya | ihrachyshka: since the consequence of breaking ci systems is they all rush in here to ask us what to do | 16:21 |
*** emagana has quit IRC | 16:21 | |
anteaya | ihrachyshka: so timing the thundering herd is helpful for us | 16:21 |
clarkb | it is weird that other projects would rely on the oslo incubator logging in neutron to do their logging | 16:21 |
clarkb | that is a bug in the split out code imo | 16:21 |
clarkb | so overall the move is good but ya communication is nice | 16:22 |
fungi | ihrachyshka: anteaya: probably the "no idea how to communicate with third-party ci operators" situation could be improved by better socializing the third-party testing irc meetings and announcements ml | 16:22 |
jeblair | fungi, clarkb, zaro: after reading the bugs and irc log from last year, i think we should convert review-dev to utf8_bin and test there. | 16:23 |
fungi | and then incorporate that switch into the 2.9 upgrade path? | 16:23 |
fungi | (potentially) | 16:23 |
jeblair | fungi: yeah, assuming it looks good | 16:23 |
jeblair | fungi: or perhaps into the os upgrade path | 16:24 |
jeblair | either one, really. | 16:24 |
*** hodos has joined #openstack-infra | 16:24 | |
*** sabeen has joined #openstack-infra | 16:24 | |
jeblair | zaro: are you up for doing that conversion on review-dev? | 16:24 |
zaro | sure, can do. | 16:25 |
fungi | sounds like a good idea to me | 16:25 |
openstackgerrit | Chris Dent proposed openstack/requirements: Update gabbi to 0.11.0 https://review.openstack.org/156253 | 16:25 |
jeblair | https://review-dev.openstack.org/#/c/5258/ | 16:25 |
*** ildikov has joined #openstack-infra | 16:25 | |
jeblair | fungi, zaro: just tested that i see '?' with review dev too | 16:25 |
*** koolhead17 has joined #openstack-infra | 16:26 | |
*** koolhead17 has quit IRC | 16:26 | |
*** koolhead17 has joined #openstack-infra | 16:26 | |
zaro | yep, will upgrade and test shortly. | 16:26 |
ihrachyshka | anteaya, I am all for extending the period, or better communication. just curious: should oslo team also consult with infra when releasing something that will break projects that rely on private symbols? | 16:27 |
ihrachyshka | anteaya, also, CI logs already provide clear problem about what failed (they clearly state missing import is the culprit) | 16:27 |
*** scheuran has quit IRC | 16:28 | |
*** ominakov_ has joined #openstack-infra | 16:28 | |
*** dizquierdo has quit IRC | 16:28 | |
fungi | ihrachyshka: anteaya: i'm less concerned about projects giving us a heads up if they're going to break someone else's ci, but that's because i'm in favor of training third-party ci operators not to come asking the infra team for help figuring out their bugs | 16:29 |
*** tnovacik has joined #openstack-infra | 16:29 | |
*** e0ne is now known as e0ne_ | 16:29 | |
fungi | where "training" them usually involves me saying "i have no idea, i'm not the one running your system, but you can look at how we run ours" | 16:30 |
*** sergsh has joined #openstack-infra | 16:30 | |
*** tiswanso has quit IRC | 16:30 | |
ihrachyshka | neutron CI world is quite bumpy this cycle, I think 3party operators are now used to the fact that something may break. L should be a bit better in this regards. | 16:30 |
*** dboik_ has joined #openstack-infra | 16:30 | |
*** tiswanso has joined #openstack-infra | 16:31 | |
*** e0ne_ is now known as e0ne | 16:31 | |
*** ayoung is now known as ayoungbot | 16:32 | |
dhellmann | fungi, sdague, jeblair, clarkb: if you have a few minutes today, this is a really small oslo-supporting change to devstack-gate: https://review.openstack.org/#/c/162218/1 | 16:32 |
fungi | ihrachyshka: but anyway, good ways to give them a heads up (because many of them are not really indoctrinated into the free software habit of monitoring high-traffic dev mailing lists) is a post to the third-party-announce ml and/or adding something to the https://wiki.openstack.org/wiki/Meetings/ThirdParty agenda | 16:32 |
jeblair | dhellmann: that's straightforward. +2 :) | 16:32 |
hodos | hi everyone can somebody help with the CI setup. I have a question about running nodepool. What's the prepare script should i use for image (prepare_devstack.sh ??? ) and where to take the base ubuntu image | 16:32 |
dhellmann | jeblair: thanks | 16:32 |
clarkb | dhellmann: that was easy approved | 16:32 |
dhellmann | clarkb: thanks | 16:32 |
ihrachyshka | fungi, oh, I didn't know about their ml, I'll cross post there then! | 16:32 |
*** ayoungbot is now known as ayoung | 16:33 | |
*** dboik has quit IRC | 16:33 | |
clarkb | hodos: you can see our config at https://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/templates/nodepool/nodepool.yaml.erb | 16:34 |
*** baoli has quit IRC | 16:34 | |
*** unicell has joined #openstack-infra | 16:34 | |
clarkb | hodos: we use prepare_node_devstack.sh according to that and the specified cloud specific base image | 16:34 |
ihrachyshka | fungi, forwarded the email into the ML | 16:35 |
hodos | aha, thank you. | 16:35 |
hodos | so how do you create that image, i just took ubuntu_server iso but it doesn't seem to work for me | 16:36 |
*** baoli_ has joined #openstack-infra | 16:36 | |
hodos | i mean i uploaded it to glance | 16:36 |
clarkb | hodos: canonical publishes ubuntu cloud images. hpcloud uses those but rackspace builds their own | 16:36 |
ihrachyshka | fungi, the email is stuck in moderation queue, so not sure it will reach people in timely manner | 16:36 |
hodos | got it. thanks | 16:37 |
jogo | mtreinish: nice, although docs directories sometimes do have python in them | 16:37 |
*** bhuvan has joined #openstack-infra | 16:37 | |
anteaya | ihrachyshka: so better communication | 16:38 |
anteaya | ihrachyshka: I have yet to see anyone ask a question in infra where we tell them you shouldn't ask that question | 16:38 |
*** pblaho_ is now known as pblaho | 16:38 | |
jogo | anteaya: yup last comment does just that https://github.com/jogo/lastcomment/blob/master/lastcomment.py | 16:38 |
*** jlanoux has quit IRC | 16:38 | |
anteaya | now we might tell them that we don't know and give them a suggestion how to help themselves | 16:38 |
fungi | speaking of cloud images, there are actually debian testing (jessie) images being published at http://cdimage.debian.org/cdimage/openstack/testing/ | 16:39 |
anteaya | ihrachyshka: but asking is always a good first step | 16:39 |
*** yamahata has quit IRC | 16:39 | |
anteaya | ihrachyshka: and I'm one of the administers of the third party announce mailing list | 16:39 |
anteaya | ihrachyshka: which is why I hope in future you will come and talk to us about things in the planning stage | 16:39 |
jogo | anteaya: as for the voting element, I totally understand where you are coming from. I am just trying to take the big tent related things to the logical continuation. So that issues like this surface and can be dealt with | 16:40 |
anteaya | ihrachyshka: so that if we have some feedback on the plans we can share | 16:40 |
*** bdpayne has joined #openstack-infra | 16:40 | |
anteaya | jogo: yeah I understand | 16:40 |
anteaya | jogo: I had no idea that you wanted to do anything other than help someone who asked | 16:40 |
ihrachyshka | anteaya, ack, will do | 16:40 |
anteaya | since righ tnow our tooling doesn't scale for the big tent | 16:40 |
anteaya | so adding a link doesn't help that | 16:40 |
*** jaypipes has joined #openstack-infra | 16:40 | |
jogo | anteaya: well two people essentially asked so wanted a generic answer | 16:40 |
jogo | anteaya: but yeah you raise a good point, we need Elections as a Service :) | 16:41 |
anteaya | so you can share that with those two people | 16:41 |
anteaya | you are welcome to particiapte in the election tooling conversation | 16:41 |
anteaya | it is my expectation it is on today's agenda for infra, I haven't checked yet thought | 16:41 |
anteaya | jogo: when you tell them be sure to tell them if they use that process to avoid the upcoming ptl tc elections | 16:42 |
*** tnovacik has quit IRC | 16:42 | |
anteaya | since having competing elections on the list at the same time confuses the electorate | 16:42 |
anteaya | ihrachyshka: thank you | 16:42 |
*** bhuvan has quit IRC | 16:43 | |
*** harlowja_away is now known as harlowja_ | 16:44 | |
*** dimsum__ has quit IRC | 16:46 | |
*** dimsum__ has joined #openstack-infra | 16:47 | |
*** ihrachyshka has quit IRC | 16:48 | |
*** Somay has quit IRC | 16:48 | |
zaro | pelix: ping | 16:49 |
BobBall | asselin, jeblair: Where's the best place to raise zuul bugs? Openstack-Gate? I don't see a zuul project in bugs.launchpad.net | 16:49 |
jogo | anteaya: FYI these are the two patches in question https://review.openstack.org/#/c/162745/ https://review.openstack.org/#/c/161902/ I haven't explicitly shared the link with them yet though | 16:49 |
jeblair | BobBall: https://storyboard.openstack.org/#!/project/679 | 16:49 |
*** andreykurilin_ has joined #openstack-infra | 16:50 | |
*** sarob has quit IRC | 16:50 | |
BobBall | Ah - ok | 16:50 |
jeblair | BobBall: and you can always find the answer to questions like that for any of our systems at http://ci.openstack.org/zuul.html | 16:50 |
BobBall | sorry - I just assumed (my bad, I know) that all the infra components were also using launchpad | 16:50 |
anteaya | jeblair: okay SergeyLukjanov is listed as one of the two murano ptl's | 16:51 |
jeblair | BobBall: we're dogfooding storyboard before we consider moving the rest of the projects to it | 16:51 |
*** amotoki has quit IRC | 16:51 | |
anteaya | jeblair: sorry was for jogo | 16:51 |
anteaya | jeblair: okay SergeyLukjanov is listed as one of the two murano ptl's | 16:51 |
*** amotoki has joined #openstack-infra | 16:51 | |
*** amotoki has quit IRC | 16:51 | |
anteaya | I did it again | 16:51 |
anteaya | jogo: for you | 16:51 |
BobBall | understood. | 16:51 |
anteaya | jogo: so SergeyLukjanov knows how to run an election | 16:51 |
*** marun has quit IRC | 16:52 | |
*** sarob has joined #openstack-infra | 16:52 | |
anteaya | and group based policy also has a ptl | 16:52 |
jogo | anteaya: but not elected | 16:52 |
anteaya | all they have to do is set up a poll | 16:53 |
anteaya | part of me is concerned with the need for them to advertise in order to tout their patch | 16:53 |
*** hyakuhei has joined #openstack-infra | 16:53 | |
anteaya | rather than actually having a crisis of leadership | 16:53 |
anteaya | so if they don't actually have a leadership crisis | 16:53 |
jogo | anteaya: yeah I hear that, but there should be some sort of election for formality no? | 16:54 |
anteaya | perhaps they can wait until we have an election workflow we can offer for thousands | 16:54 |
anteaya | well formality costs | 16:54 |
openstackgerrit | James E. Blair proposed stackforge/gertty: Hide webbrowser output https://review.openstack.org/163085 | 16:54 |
anteaya | right now at the very least in terms of noise on the mailing list | 16:54 |
fungi | jogo: even for official projects we only have an election if there's more than one nominee | 16:54 |
jogo | The leadership is chosen by the contributors to the project | 16:54 |
*** hyakuhei has quit IRC | 16:54 | |
anteaya | which I would like to keep clear for the ptl tc elections | 16:55 |
openstackgerrit | Doug Hellmann proposed openstack/requirements: Cap Oslo libraries for kilo https://review.openstack.org/162656 | 16:55 |
anteaya | jogo: I can't stop them | 16:55 |
anteaya | but I'm not a fan | 16:55 |
anteaya | not right now | 16:55 |
jeblair | anteaya: when does election season start? | 16:55 |
jogo | anteaya: so ignoring the very real conflict with upcoming elections for a second | 16:55 |
pelix | zaro: pong | 16:56 |
jogo | anteaya: what do you think the process should be a for a new project to meet the following req | 16:56 |
jogo | 'The leadership is chosen by the contributors to the project' | 16:56 |
*** esker has joined #openstack-infra | 16:56 | |
anteaya | jeblair: well right now tristanC and I selected dates in april | 16:57 |
anteaya | but we haven't announced them for a few reasons | 16:57 |
anteaya | one: election tooling | 16:57 |
anteaya | two: electoral roll creaetion including foundation db access and atc definition | 16:57 |
*** marun has joined #openstack-infra | 16:58 | |
*** agentle_ has quit IRC | 16:58 | |
*** bhuvan has joined #openstack-infra | 16:58 | |
clarkb | jenkins02 is starting back up now | 16:59 |
zaro | pelix: take a look at the comments https://review.openstack.org/#/c/162045/ | 16:59 |
zaro | pelix: opps meant this one: https://review.openstack.org/#/c/156793/ | 16:59 |
anteaya | jogo: the leadership is chosen by the contributors to the project, can take any form | 17:00 |
*** gyee has quit IRC | 17:00 | |
openstackgerrit | Terry Wilson proposed openstack/requirements: Bump oslo.rootwrap to 1.6.0 https://review.openstack.org/163089 | 17:00 |
pelix | zaro: phew! I was getting worried there for second | 17:00 |
anteaya | as long as the contributors feel included | 17:00 |
zaro | pelix: we have an char encoding problem. was wondering if you have the same issue with your gerrit? | 17:00 |
jogo | anteaya: agreed, do you have any forms you would recommend? | 17:00 |
anteaya | jogo: the clue here would be any contributor saying "hey, I wasn't included" | 17:00 |
anteaya | jogo: they can vote during a meeting | 17:00 |
jogo | anteaya: presumably something that has a paper trail showing there was an election? | 17:00 |
anteaya | meetings are logged and the meetbot does voting well | 17:01 |
anteaya | jogo: that is what I think we don't need yet | 17:01 |
jogo | anteaya: that could work, except for folks who couldn't attent | 17:01 |
*** jistr has quit IRC | 17:01 | |
anteaya | meeting logs are enough of a paper trail for me | 17:01 |
zaro | pelix: noticed your name in this issue: https://bugs.launchpad.net/openstack-ci/+bug/979227 | 17:01 |
openstack | Launchpad bug 979227 in OpenStack Core Infrastructure "convert gerrit's mysql tables to utf8" [Medium,Fix released] - Assigned to Khai Do (zaro0508) | 17:01 |
pelix | zaro: don't think so, but then I converted us to using utf8 and innodb a while back | 17:01 |
anteaya | jogo: then that is the project's problem | 17:01 |
jogo | although to be honest, I doubt there will be more then one candidate in most cases | 17:01 |
anteaya | jogo: besides appearing legitmate, what leadership issue aer we trying to solve today? | 17:01 |
zaro | pelix: ok. we were going to do the same, just wanted to see if you are experience the same issue so i don't have to go thru the trouble. | 17:02 |
*** gyee has joined #openstack-infra | 17:02 | |
anteaya | jogo: precisely why I would like to defer this discussion until after we have tooling for the masses | 17:02 |
anteaya | jogo: if a project legitmiately has a crisis in leadership they can ask for guidance and support | 17:02 |
anteaya | and I will do what I can | 17:02 |
*** Ryan_Lane has joined #openstack-infra | 17:02 | |
anteaya | so far the need is to market their patch | 17:02 |
jogo | anteaya: simply the appearing legitimate issue, I don't think we want to wait for things to have a crisas | 17:02 |
anteaya | to governance | 17:03 |
pelix | zaro: well I converted us not because of the charset issue, but because at the time needed to use utf8 and innodb for the search function in gerrit to work correctly | 17:03 |
*** yamahata has joined #openstack-infra | 17:03 | |
anteaya | jogo: like I said, today I have nothing to offer | 17:03 |
anteaya | except I can't stop them | 17:03 |
anteaya | ask me after the infra meeting | 17:03 |
anteaya | which takes place in 2 hours | 17:03 |
anteaya | perhaps I might have something then | 17:03 |
jogo | anteaya: sounds like a plan | 17:03 |
pelix | but that got solved for MyISAM or at least wasn't a problem subsequently for openstack's gerrit so I think it got shelved | 17:03 |
jogo | anteaya: I don't have any good answers myself | 17:03 |
*** banix has joined #openstack-infra | 17:04 | |
anteaya | well here we sit, trying to solve a problem that doesn't exist | 17:04 |
*** sbadia has quit IRC | 17:04 | |
anteaya | exhausting ourselves from focusing on the problem that actually does | 17:04 |
*** sbadia has joined #openstack-infra | 17:04 | |
anteaya | which is do we have options for election tooling | 17:04 |
anteaya | I welcome your thoughts on that | 17:04 |
pelix | zaro: only thing I do recall is you need to tweak the connection string in gerrit when you switch in order to avoid issues in the future with DB upgrades | 17:05 |
jogo | anteaya: I actually think this problem exists, having a recommended way to show the paper trail from some sort of process that shows openness is important | 17:05 |
*** ivar-lazzaro has joined #openstack-infra | 17:05 | |
anteaya | I think a bigger problem exists | 17:05 |
jogo | anteaya: even if the answer is, you can simply do it in your weekly meeting | 17:05 |
anteaya | did you not hear anything I said? | 17:06 |
*** changbl has joined #openstack-infra | 17:06 | |
anteaya | is that the only thing you heard me say? | 17:06 |
*** Longgeek has quit IRC | 17:06 | |
anteaya | I said several times we are in the middle of discussing how to solve our election tooling scaling issue | 17:06 |
anteaya | didn't you hear me say that? | 17:06 |
pelix | zaro: had to set database.url in the gerrit config file to: jdbc:mysql://localhost/reviewdb?useUnicode=yes&characterEncoding=UTF-8&sessionVariables=storage_engine=InnoDB | 17:06 |
pelix | think that avoids any issues with what the system defaults might be | 17:07 |
*** gyee has quit IRC | 17:07 | |
*** jaypipes has quit IRC | 17:07 | |
fungi | pelix: that sounds useful | 17:07 |
zaro | pelix: so you didn't convert the db at all? just added that to the conn string? | 17:07 |
jogo | I did and I understand, just pointing out that I think this is a problem that does exist (although there are definitely other bigger issues to tackle around this first) | 17:08 |
*** markvoelker has joined #openstack-infra | 17:09 | |
*** yamamoto has joined #openstack-infra | 17:09 | |
anteaya | did I say it didn't exist? | 17:10 |
anteaya | and yes you pointed it several times | 17:11 |
anteaya | those shopping for legitmacy for their goverance patches have a gap | 17:11 |
anteaya | got it | 17:11 |
anteaya | what they don't have is a leadership crisis | 17:11 |
anteaya | which I would address today if it arose | 17:11 |
jogo | anteaya: ahh, that makes more sense | 17:12 |
jogo | anteaya: sounds like we are in agreement then | 17:12 |
*** yfried_ has joined #openstack-infra | 17:12 | |
jogo | anteaya: thanks! | 17:12 |
anteaya | some time ago | 17:12 |
jogo | anteaya: :) | 17:12 |
anteaya | thanks | 17:12 |
anteaya | I appreciate your zeal | 17:12 |
fungi | one of the things i think we discussed as a requirement for entry into the governance reference/projects.yaml file, though i'm unsure whether it made it into the current design, is to note the model by which a project expects to identify its electorate | 17:12 |
fungi | (for ptl elections) | 17:13 |
jogo | fungi: oh interesting, I didn't see any thing in the governance repo saying that today (may have missed it though) | 17:13 |
anteaya | fungi: good point, and another fish on the table to kick around as we figure out how to scale this | 17:13 |
*** pblaho_ has joined #openstack-infra | 17:13 | |
fungi | we could extend that to document the model by which that electorate chooses a their ptl | 17:13 |
anteaya | there was a lot of governance stuff that didn't make it into the current model | 17:14 |
*** yamamoto has quit IRC | 17:14 | |
fungi | jogo: that was more to support non-code projects, for example translations might hold elections based on participants polled from zanata | 17:14 |
anteaya | as I was told that would be decided by a working group once the big stuff was figured out | 17:14 |
jeblair | clarkb: you had a possible zuul bug? | 17:14 |
*** krtaylor has quit IRC | 17:14 | |
*** sergsh has quit IRC | 17:15 | |
clarkb | jeblair: yes. change 162887 was reported as a merge failure at 1240UTC by zuul to gerrit. However a recheck kicked off the jobs for that change without a rebase | 17:15 |
clarkb | jeblair: looking in the zuul debug log at about 1146UTC there is a merge complete event for that change with state PENDING | 17:15 |
*** unicell has quit IRC | 17:15 | |
*** Somay has joined #openstack-infra | 17:15 | |
*** unicell has joined #openstack-infra | 17:15 | |
clarkb | it almost appears that if we get a merge event in a PENDING state that prepareRef and _processOneItem() in the zuul scheduler don't handle that properly | 17:16 |
clarkb | but I am still reading through that portion of the code to see if I can figure that out | 17:16 |
*** pblaho has quit IRC | 17:16 | |
clarkb | jeblair: I think that onMergeCompleted or the event being placed to fire it are the root cause because onMergeCompleted does not distinguish based on state | 17:18 |
clarkb | so either it needs to distinguish or we need to not set an event that will fire that handler when state is still PENDING | 17:18 |
*** melwitt has joined #openstack-infra | 17:18 | |
*** markvoelker has quit IRC | 17:19 | |
jeblair | clarkb: if the merge is complete, the state will not be pending for much longer. | 17:19 |
*** achanda has joined #openstack-infra | 17:19 | |
*** achanda has quit IRC | 17:19 | |
jeblair | clarkb: pending means waiting for a response from the merger | 17:19 |
*** markvoelker has joined #openstack-infra | 17:19 | |
*** achanda has joined #openstack-infra | 17:19 | |
*** AJaeger_ has joined #openstack-infra | 17:20 | |
*** AJaeger_ has quit IRC | 17:20 | |
*** yfried_ has quit IRC | 17:21 | |
*** sdake_ has quit IRC | 17:21 | |
*** yfried_ has joined #openstack-infra | 17:21 | |
*** emagana has joined #openstack-infra | 17:21 | |
*** changbl has quit IRC | 17:21 | |
*** markvoelker_ has joined #openstack-infra | 17:21 | |
clarkb | jeblair: I am unable to find where the zm0X hosts actually ran the merge though which makes me think it was still actually pending | 17:21 |
clarkb | I checked 01-08 | 17:21 |
*** patrickeast has joined #openstack-infra | 17:22 | |
zaro | jeblair, clarkb, fungi : thanks to pelix, i think the problem is fixed https://review-dev.openstack.org/#/c/5240 | 17:23 |
*** markvoelker has quit IRC | 17:24 | |
zaro | didn't need to run the conversion. | 17:24 |
clarkb | jeblair: maybe it is a race between _processOneItem() calling prepareRef() and the pipeline manager getting the onMergeCompleted event? | 17:24 |
*** ddieterly has quit IRC | 17:24 | |
*** changbl has joined #openstack-infra | 17:24 | |
zaro | will create a change for it. | 17:25 |
*** hdd_ has quit IRC | 17:25 | |
*** devvesa has quit IRC | 17:25 | |
dougwig | hi infra. would it be possible to dedicate a node or two to the experimental queue? it seems stuck, and makes testing new jobs rather painful. :) | 17:25 |
anteaya | dougwig: we don't have that ability | 17:26 |
dougwig | is there an alternative, like non-voting in the check queue, or bumping it's priority? the queue is short, and basically dead. | 17:26 |
*** MaxV has quit IRC | 17:27 | |
*** krtaylor has joined #openstack-infra | 17:27 | |
anteaya | seems like both rax and hp are having high error rates at the moment | 17:27 |
mestery | yikes | 17:27 |
jeblair | we could set the experimental queue priority to be equal to check | 17:27 |
* mestery is tiliting at the experimental queue as well | 17:27 | |
marun | jeblair: that would be much appreciated. the alternative is we waste nodes running non-voting in the check queue | 17:27 |
clarkb | jeblair: it appears that gearman is async there so it may indeed race since it short circuits the next time through if it is still pending | 17:28 |
dougwig | jeblair: that'd be preferable to me to incurring my test jobs on everyone's patches. | 17:28 |
*** hdd_ has joined #openstack-infra | 17:28 | |
fungi | clarkb: i think my login.defs change did at least fix the race i was originally seeing http://paste.openstack.org/show/191356/ | 17:28 |
jeblair | clarkb: i'm having trouble following your logic. can you etherpad relevant log lines? | 17:28 |
clarkb | jeblair: yes | 17:28 |
marun | jeblair: should we propose a patch? | 17:29 |
jeblair | marun: please | 17:29 |
marun | jeblair: will do, thank you. | 17:29 |
dougwig | marun: are you on that, or do you want me to get it? | 17:30 |
*** MaxV has joined #openstack-infra | 17:30 | |
marun | dougwig: on it | 17:30 |
dougwig | marun: thanks. | 17:30 |
clarkb | jeblair: https://etherpad.openstack.org/p/debug-zuul-pending-merges | 17:30 |
*** johnthetubaguy is now known as zz_johnthetubagu | 17:31 | |
pelix | zaro: also converted the DB, the connection string was to ensure that subsequent upgrades of gerrit (which can change the DB schema) wouldn't accidentally use the system default charset or engine type for new tables added | 17:31 |
*** MaxV has quit IRC | 17:31 | |
*** stevemar has quit IRC | 17:32 | |
*** coolsvap|afk is now known as coolsvap | 17:32 | |
*** changbl has quit IRC | 17:33 | |
zaro | pelix: hmm, well just adding that conn string seems to have fixed it for us. | 17:34 |
*** unicell has quit IRC | 17:34 | |
*** bhunter71 has quit IRC | 17:34 | |
*** unicell has joined #openstack-infra | 17:34 | |
*** achanda has quit IRC | 17:35 | |
*** ominakov_ has quit IRC | 17:35 | |
*** tnovacik has joined #openstack-infra | 17:36 | |
openstackgerrit | Maru Newby proposed openstack-infra/project-config: Set experimental pipeline precedence to 'normal' https://review.openstack.org/163101 | 17:36 |
marun | dougwig: ^^ | 17:36 |
pelix | zaro: might be the case that your system default charset was already utf-8 so in your case the only problem was the connection not using the correct charset. | 17:36 |
fungi | zaro: jeblair: yep, now i get kanji and katakana just fine in comments at https://review-dev.openstack.org/5258 | 17:36 |
*** achanda has joined #openstack-infra | 17:37 | |
openstackgerrit | Maru Newby proposed openstack-infra/project-config: Set experimental pipeline precedence to 'normal' https://review.openstack.org/163101 | 17:37 |
*** jp_at_hp has quit IRC | 17:38 | |
anteaya | zaro fungi nicely done | 17:38 |
* fungi did nothing | 17:39 | |
anteaya | you are good at that too | 17:39 |
fungi | i have lots of practice | 17:39 |
anteaya | I hear that is what it takes | 17:39 |
*** _nadya_ has quit IRC | 17:40 | |
*** esker has quit IRC | 17:40 | |
*** timcline has joined #openstack-infra | 17:40 | |
*** changbl has joined #openstack-infra | 17:41 | |
*** e0ne has quit IRC | 17:42 | |
*** yfried_ is now known as yfried|afk | 17:42 | |
openstackgerrit | Khai Do proposed openstack-infra/puppet-gerrit: Fix Gerrit character encoding https://review.openstack.org/163104 | 17:42 |
zaro | clarkb, fungi, jeblair, pelix ^ | 17:42 |
* pelix looks at fungi closely to see if he resembles Wally | 17:43 | |
pelix | :p | 17:43 |
*** HeOS has quit IRC | 17:43 | |
jeblair | clarkb: assuming the merge complete log line you found is the matching one (unfortunately, the log lines are not constructed well enough to confirm that, i think), that could have been caused by a work_fail or work_exception packet | 17:45 |
jeblair | clarkb: or disconnect | 17:45 |
clarkb | jeblair: I see | 17:46 |
*** yfried|afk is now known as yfried_ | 17:46 | |
yolanda | hi, does infra meeting time change with dailylight saving changes? or is it still at same UTC hour? | 17:47 |
jeblair | yolanda: same utc time | 17:47 |
fungi | yolanda: always 1900 utc | 17:47 |
yolanda | ok thx | 17:48 |
fungi | same bat time, same bat channel | 17:48 |
*** dprince has quit IRC | 17:48 | |
*** Somay has quit IRC | 17:48 | |
clarkb | jeblair: the mergers have not restarted so any disconnects would've happened without the daemons dying | 17:48 |
yolanda | i was planning to attend today, thx for the info | 17:48 |
*** bhunter71 has joined #openstack-infra | 17:49 | |
*** otter768 has joined #openstack-infra | 17:50 | |
*** esker has joined #openstack-infra | 17:51 | |
*** ChuckC has joined #openstack-infra | 17:51 | |
crinkle | mordred: the packaging session might be relevant to your interests | 17:52 |
crinkle | question came up, "why doesn't infra do this?" | 17:52 |
anteaya | zaro: you have the best commit message titles: my mama says to test closes bug | 17:52 |
*** peristeri has joined #openstack-infra | 17:53 | |
anteaya | zaro: what effect with that have on current data? | 17:53 |
mordred | crinkle: oh, be right there | 17:54 |
anteaya | fungi: I have my phone on utc time, I love it since I don't have to change my meeting schedule pattern | 17:54 |
*** weshay has quit IRC | 17:54 | |
zaro | anteaya: need to check with my mama | 17:54 |
fungi | anteaya: me too! | 17:54 |
anteaya | zaro: ha ha ha | 17:54 |
*** resker has joined #openstack-infra | 17:54 | |
*** otter768 has quit IRC | 17:55 | |
anteaya | zaro: I'll add my question to the patch | 17:55 |
clarkb | crinkle what is "this"? | 17:55 |
*** craigbr has quit IRC | 17:55 | |
*** tiswanso has quit IRC | 17:55 | |
fungi | ops meetup session | 17:55 |
crinkle | clarkb: building packages | 17:55 |
clarkb | we have talked about packages for our testing but I think we dont want to be in business of building packages for deployment | 17:56 |
clarkb | that is what distros do | 17:56 |
*** markus_z has quit IRC | 17:56 | |
*** tiswanso has joined #openstack-infra | 17:56 | |
*** weshay has joined #openstack-infra | 17:56 | |
*** yfried_ is now known as yfried|afk | 17:56 | |
fungi | but also, the general answer to "why doesn't infra do <x> for us" is that we help projects help themselves, and nobody has stepped up to do something like that | 17:56 |
clarkb | +1 yall know where to find us :) | 17:57 |
*** esker has quit IRC | 17:57 | |
openstackgerrit | James E. Blair proposed openstack-infra/zuul: Improve merge client logging https://review.openstack.org/163110 | 17:57 |
*** melwitt_ has joined #openstack-infra | 17:58 | |
*** andreykurilin_ has quit IRC | 17:58 | |
*** melwitt has quit IRC | 17:58 | |
*** ihrachyshka has joined #openstack-infra | 18:00 | |
anteaya | but my latest foray shows me many folks don't want to bug us | 18:00 |
anteaya | as they see us as too busy | 18:00 |
anteaya | like keystone and cinder | 18:00 |
fungi | that's a circular argument | 18:00 |
*** asselin_ has joined #openstack-infra | 18:00 | |
clarkb | if jobs were proposed to build and publish packages with people interested in taking care of them we would run those jobs | 18:00 |
*** Somay has joined #openstack-infra | 18:01 | |
anteaya | regardless of what kind of argument it is | 18:01 |
anteaya | the perception exists | 18:01 |
anteaya | and I get many basic how does infra work questions at mid-cycles | 18:01 |
fungi | helping us get stuff done isn't necessarily bugging us or adding to our workload, and potentially leads down the path to making us less busy | 18:01 |
anteaya | oh _I_ agree | 18:01 |
anteaya | but they just see the channel always full | 18:02 |
anteaya | and do something else | 18:02 |
fungi | however, a reciprocal investment is a bit of a commitment and _they_ may be too busy | 18:02 |
greghaynes | waynr: Heyo, something I forgot to mention - I did make some small amount of progress in implementing the github webhook listener as a connection driver | 18:02 |
anteaya | that also plays a role, yes | 18:03 |
greghaynes | waynr: which might be relevant to your spec writing | 18:03 |
fungi | just like those projects don't really want contributors dropping drive-by feature adds on them, we're in basically the same boat | 18:03 |
*** jaypipes has joined #openstack-infra | 18:04 | |
*** tsg_ has quit IRC | 18:04 | |
*** baoli_ has quit IRC | 18:05 | |
*** changbl has quit IRC | 18:05 | |
*** craigbr has joined #openstack-infra | 18:06 | |
*** baoli has joined #openstack-infra | 18:07 | |
*** dteselkin has quit IRC | 18:07 | |
*** tsufiev_ has quit IRC | 18:07 | |
*** katyafervent has quit IRC | 18:07 | |
*** rakhmerov has quit IRC | 18:07 | |
*** freerunner has joined #openstack-infra | 18:08 | |
*** krisgan has quit IRC | 18:08 | |
cinerama | i'm confused - are we meeting now? | 18:08 |
anteaya | 1900 | 18:08 |
anteaya | it is 1808 | 18:08 |
anteaya | keystone meeting | 18:08 |
jeblair | fungi: let's talk about https://review.openstack.org/#/c/158371/ | 18:08 |
waynr | greghaynes: cool, could you make that WIP available for me to reference? | 18:09 |
cinerama | ah ok. some of the time conversion sites are just plain wrong | 18:09 |
greghaynes | waynr: yep | 18:09 |
fungi | jeblair: sure! should we be putting more details on the wiki than we do now? | 18:09 |
jeblair | fungi: i agree with everything you wrote; however, i feel weird with us being the odd project out | 18:09 |
anteaya | cinerama: is it a converstaion site we have linked to in our documentation? | 18:09 |
jeblair | fungi: since all the other projects point to the wiki | 18:09 |
*** katyafervent has joined #openstack-infra | 18:09 | |
anteaya | cinerama: is there a problem with our tools? | 18:09 |
cinerama | anteaya: nope, was just being lazy instead of using "date" :) | 18:10 |
*** katyafervent has quit IRC | 18:10 | |
fungi | jeblair: i'm all for consistency where it makes sense. however there's also a drive to move more authoritative reference material out of the wiki | 18:10 |
*** EntropyWorks has joined #openstack-infra | 18:10 | |
anteaya | cinerama: ah okay | 18:10 |
fungi | jeblair: but i'm willing to buy the argument that we should move everyone at once if there's a suitable mechanism and location for hosting that | 18:10 |
fungi | and until there is, just keep a stub for infra in the wiki (or put more in the wiki than we currently do) | 18:11 |
openstackgerrit | greghaynes proposed openstack-infra/zuul: WIP: Support GitHub PR webhooks https://review.openstack.org/163117 | 18:11 |
jeblair | fungi: i wouldn't necessarily make the argument based entirely on consistency.... | 18:11 |
greghaynes | waynr: ^ | 18:11 |
jeblair | fungi: and if other projects move, i don't know that there's a need for them to all move at once | 18:11 |
greghaynes | very WIP but there you go :) | 18:12 |
jeblair | fungi: but if you look at https://wiki.openstack.org/wiki/Nova | 18:12 |
fungi | jeblair: assumed the reason for a url in the yaml file rather than just a wiki article name was to facilitate piecemeal relocation | 18:12 |
openstackgerrit | Kiall Mac Innes proposed openstack-infra/project-config: Enable Tempest in Designate dsvm jobs https://review.openstack.org/163118 | 18:12 |
jeblair | fungi: there's a lot of stuff in there that isn't (and probably shouldn't) be in the governance repo | 18:12 |
*** shardy has quit IRC | 18:13 | |
fungi | jeblair: agreed. a lot of the things they have there are analogous to what we have on ci.openstack.org | 18:13 |
*** katyafervent has joined #openstack-infra | 18:13 | |
openstackgerrit | Kiall Mac Innes proposed openstack-infra/project-config: Enable Tempest in Designate dsvm jobs https://review.openstack.org/163118 | 18:13 |
*** peristeri has quit IRC | 18:14 | |
*** EntropyWorks has left #openstack-infra | 18:14 | |
*** rakhmerov has joined #openstack-infra | 18:14 | |
openstackgerrit | Matthias Runge proposed openstack/requirements: Raise cap for Django https://review.openstack.org/155353 | 18:14 |
jeblair | fungi: at any rate, i think the change seems reasonable, and maybe we should go ahead with it, even if other projects do not do something similar. | 18:14 |
waynr | greghaynes: so it looks like you weren't planning to use an existing github api wrapper? | 18:14 |
jeblair | fungi: because you are right, it's pretty duplicative. we should probably keep that wiki page though as a redirect. | 18:15 |
*** mrunge has quit IRC | 18:15 | |
*** dteselkin has joined #openstack-infra | 18:15 | |
greghaynes | waynr: eh, we could and id be +1 on it, I mostly was trying to get the driver + server figured out | 18:15 |
*** sarob has quit IRC | 18:15 | |
clarkb | waynr: greghaynes you mean some proxy service? | 18:16 |
*** tsufiev_ has joined #openstack-infra | 18:16 | |
greghaynes | and enough info in it that I coul dget a minimal test working | 18:16 |
jeblair | fungi: mostly i wanted to sync up on whether you knew of a widespread effort to do something similar, and if not, whether we still felt okay being the only project that did not link to the wiki there. | 18:16 |
clarkb | or eg PyGithub? | 18:16 |
greghaynes | clarkb: im thinking the latter | 18:16 |
fungi | jeblair: in some ways i was hoping it might serve as an example that the wiki is not necessarily a canonical reference for that sort of stuff, and that projects are free to maintain those details in other ways/places as suits their workflow | 18:16 |
*** yfried|afk is now known as yfried_ | 18:16 | |
jeblair | fungi: works for me. if you're still +1 i'm +1. | 18:17 |
waynr | clarkb: greghaynes i was thinking github3.py (by sigmavirus24) | 18:17 |
greghaynes | waynr: Yep, I really have no strong opinions either way on that, I thing parsing the response data is the easy bit of work | 18:17 |
*** gyee has joined #openstack-infra | 18:17 | |
fungi | jeblair: if nothing else, it gives the tc an opportunity to consider whether or not that's what they actually intend by having freeform homepage urls in the list, and engage in a discussion if not | 18:18 |
clarkb | my only suggestion there is we use pygithub for other things so have some familiarity with it | 18:18 |
waynr | greghaynes: clarkb okay i'll take this discussion into consideration while writing up the spec, which i hope to get to today | 18:18 |
clarkb | but I also think that project is looking for a maintainer so likely not a good option (unless that has changed0 | 18:19 |
jeblair | waynr: ++ | 18:19 |
greghaynes | waynr: awesome! | 18:19 |
openstackgerrit | Sabari proposed openstack/requirements: Bump glance_store to latest version https://review.openstack.org/163121 | 18:19 |
*** dhritishikhar has quit IRC | 18:19 | |
Kiall | Infra folks, have I gone about enabling tempest install backwards here? https://review.openstack.org/#/c/163118/2/jenkins/jobs/designate.yaml or is adding tempest to ENABLED_SERVICES enough? | 18:19 |
Kiall | I'm thinking I might have :) | 18:20 |
waynr | clarkb: yeah the reason i began evaluating github3.py is that it looks to be actively maintained and even if it doesn't support all the latest github api developments, there are at least issues ticketed for addressing them | 18:20 |
clarkb | Kiall: thats the way to do it if you want tempest installed without running the tests | 18:20 |
clarkb | Kiall: DEVSTACK_GATE_TEMPEST=1 without the NOTESTS flag means install and run tempest. with the NOTESTS flag means just isntall it | 18:20 |
Kiall | And, is there a need for tempest to be included in ENABLED_SERVICES? We'll run the tests ouselves, as there in our repo and using tempest-lib + a tiny bit of tempest itself... | 18:21 |
clarkb | Kiall: no the DEVSTACK_GATE_TEMPEST=1 puts it in enabled services | 18:22 |
*** cdent has quit IRC | 18:22 | |
Kiall | Cool, Okay.. Thanks :) | 18:22 |
*** ddieterly has joined #openstack-infra | 18:22 | |
*** dizquierdo has joined #openstack-infra | 18:22 | |
Kiall | Feel free to +2 it while you have it open :P | 18:22 |
*** ssam2 has quit IRC | 18:22 | |
freerunner | Hi there! Can you help me, please? We have a patchset for Murano/Congress/Mistral integration job, but enviroment variable that we need exported incorrectly and Murano lib for devstack can't see it and set parameters by default. What we did incorrectly? https://review.openstack.org/#/c/162971 | 18:23 |
fungi | freerunner: yep, taken care of | 18:23 |
*** ddieterly has quit IRC | 18:23 | |
fungi | freerunner: out of curiosity, is that a devstack-gate environment variable? | 18:24 |
clarkb | freerunner: fungi: devstack is sudo'd dropping the env | 18:24 |
clarkb | so you have to put it in the localrc | 18:24 |
fungi | yeah, that's why i asked. it didn't look like it would be | 18:24 |
*** ddieterly has joined #openstack-infra | 18:25 | |
sigmavirus24 | clarkb: look at it like this, if y'all use github3.py I have a reason to work on it during the workday too ;) | 18:26 |
*** ghostpl_ has quit IRC | 18:26 | |
*** ddieterl_ has joined #openstack-infra | 18:26 | |
clarkb | ha | 18:26 |
Kiall | thanks clarkb :) | 18:26 |
*** yfried_ is now known as yfried|afk | 18:26 | |
freerunner | fungi: Not, this is not a devstack-gate variable | 18:26 |
sigmavirus24 | clarkb: so you know, things get fixed fasterer ;) | 18:27 |
*** dimsum__ is now known as dims | 18:27 | |
*** tnovacik has quit IRC | 18:27 | |
*** resker has quit IRC | 18:27 | |
freerunner | clarkb: So, we need to echo this variable to localrc? | 18:27 |
fungi | freerunner: right, so when you're setting environment variables in devstack-gate job definitions, those really only apply to devstack-gate itself. it needs additional logic if it's going to use that as an option to modify the localrc before starting devstack | 18:27 |
clarkb | freerunner: there is a devstack gate variable that you can set to append to the localrc, I would just set this var to include your var | 18:28 |
*** sdake_ has joined #openstack-infra | 18:28 | |
*** ddieterly has quit IRC | 18:30 | |
*** ddieterly has joined #openstack-infra | 18:30 | |
*** ddieterl_ has quit IRC | 18:31 | |
*** AJaeger_ has joined #openstack-infra | 18:32 | |
Kiall | fungi: Could I ask you to give this a quick review? 3 liner devstack-gate change About to head home, would be excllent if I could have that merged so I can get the designate side in this evening :) https://review.openstack.org/#/c/163118/ | 18:32 |
freerunner | clarkb: Uh. Could u plz tell me, what i need to do? What code should I add to patch? Could u show me any example, please? | 18:32 |
*** che-arne has quit IRC | 18:33 | |
*** jcoufal has quit IRC | 18:33 | |
Kiall | freerunner: in Designate, we do this: https://github.com/openstack/designate/blob/master/contrib/devstack/gate_hook.sh | 18:33 |
clarkb | freerunner: https://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/ec2-api.yaml#n36 | 18:33 |
Kiall | L12-13 add the vars into our localrc file so there picked up by devstack | 18:33 |
*** teran_ has quit IRC | 18:34 | |
clarkb | freerunner: that variable's contents will be appended to the localrc so just add anything you need to pass through there | 18:34 |
clarkb | freerunner: also consider updating devstack to do this instead if it should just be enabled | 18:34 |
*** e0ne has joined #openstack-infra | 18:37 | |
*** hdd_ has quit IRC | 18:41 | |
tristanC | jogo: Hi, I'm part of the election officials and I try to understand your concern... What can we do to help you ? | 18:41 |
*** stevemar has joined #openstack-infra | 18:41 | |
freerunner | clarkb: Kiall: fungi: Thank you very much ;) | 18:42 |
*** pelix has quit IRC | 18:42 | |
*** dprince has joined #openstack-infra | 18:42 | |
*** craigbr has quit IRC | 18:42 | |
tristanC | the real issue with a large number of program to run election for is: everything is done manually: confirming candidacy, wiki edition, feeding the voting tool with atc mails, ... all by hand, dual checked by both election officials | 18:43 |
openstackgerrit | Merged openstack-infra/project-config: Allow puppet-httpd to be published to the forge https://review.openstack.org/161022 | 18:43 |
*** craigbr has joined #openstack-infra | 18:43 | |
*** changbl has joined #openstack-infra | 18:44 | |
tristanC | (without mentioning the side effect of spamming the development mailing list) | 18:44 |
openstackgerrit | Victor Ryzhenkin proposed openstack-infra/project-config: Fix for gate-murano-congress-devstack-dsvm https://review.openstack.org/162971 | 18:44 |
tristanC | so my question is, are we considering handling election for stakeforge program as well ? | 18:45 |
jogo | tristanC: no, the issue here is for projects that are applying to move over to openstack | 18:45 |
jogo | we require them to elect leadership | 18:45 |
*** krotscheck has joined #openstack-infra | 18:46 | |
*** yfried|afk is now known as yfried_ | 18:46 | |
jogo | tristanC: from the governance repo 'The leadership is chosen by the contributors to the project' | 18:46 |
openstackgerrit | Merged openstack-infra/project-config: Remove use of DEVSTACK_GATE_TESTR_ARTIFACT_TARGET https://review.openstack.org/161421 | 18:46 |
jogo | tristanC: so the question is what do we tell projects that are applying | 18:46 |
*** baoli has quit IRC | 18:48 | |
tristanC | can't we tell them that until they are accepted, they'll have to run election on their own ? Using condorcet or by other means... | 18:49 |
openstackgerrit | Merged openstack-infra/project-config: Add experimental plugin based ironic+ironicclient functional jobs https://review.openstack.org/161483 | 18:49 |
anteaya | tristanC: they do have to run the election on their own | 18:49 |
anteaya | tristanC: jogo is not saying that he expects the election officials to run stackforge elections | 18:50 |
anteaya | tristanC: he is asking for some workflow to point them at | 18:50 |
anteaya | tristanC: and currently we don't have a solution that scales for all of stackforge at the same time as suiting the ptl tc elctions | 18:50 |
openstackgerrit | Paul Michali proposed openstack-infra/project-config: VPNaaS make new functional test non-voting https://review.openstack.org/163133 | 18:50 |
*** agentle_ has joined #openstack-infra | 18:50 | |
jroll | I would think stackforge projects tend to be small enough where an election via mailing list might be enough | 18:52 |
dhellmann | anteaya: if you have time today, could you take another look at https://review.openstack.org/#/c/161447 for me, please/ | 18:52 |
dhellmann | ? | 18:52 |
*** Somay has quit IRC | 18:52 | |
*** ghostpl_ has joined #openstack-infra | 18:53 | |
openstackgerrit | James E. Blair proposed openstack-infra/project-config: Fix zuul log path again https://review.openstack.org/163134 | 18:53 |
openstackgerrit | James E. Blair proposed openstack-infra/project-config: Fix propose-project-config-update job https://review.openstack.org/163135 | 18:53 |
anteaya | dhellmann: yeah sorry | 18:53 |
dhellmann | anteaya: not a big deal, I know there are other things going on :-) | 18:53 |
anteaya | dhellmann: I don't mean to ignore you | 18:53 |
*** baoli has joined #openstack-infra | 18:53 | |
anteaya | jroll: the problem is they want attention for their governance patch | 18:53 |
anteaya | jroll: it isn't a problem of leadership, it is a problem of optics | 18:54 |
*** baoli has quit IRC | 18:54 | |
jhesketh | Morning | 18:55 |
*** agentle_ has quit IRC | 18:55 | |
clarkb | jhesketh: good morning | 18:55 |
*** agentle_ has joined #openstack-infra | 18:55 | |
anteaya | morning jhesketh | 18:55 |
jroll | anteaya: I'm not sure I understand the problem, then | 18:56 |
openstackgerrit | Victor Ryzhenkin proposed openstack-infra/project-config: Fix for gate-murano-congress-devstack-dsvm https://review.openstack.org/162971 | 18:56 |
*** hdd_ has joined #openstack-infra | 18:56 | |
*** mpopow_ has joined #openstack-infra | 18:56 | |
*** claudiub has quit IRC | 18:56 | |
*** Somay has joined #openstack-infra | 18:57 | |
jogo | anteaya: correct. Any answer is adequate IMHO long as we have one that doesn't cause issues etc. | 18:57 |
*** claudiub has joined #openstack-infra | 18:57 | |
anteaya | jogo: right, which tristanC need to clarify | 18:57 |
anteaya | jroll: the problem is that openstack is made of people | 18:58 |
*** Somay has quit IRC | 18:58 | |
anteaya | who are doing good things | 18:58 |
anteaya | and people want that | 18:58 |
anteaya | so they want to have their projects be part of that | 18:58 |
anteaya | and they are checking all the boxes to do that | 18:58 |
*** craigbr has quit IRC | 18:58 | |
anteaya | in hopes of being included | 18:58 |
openstackgerrit | Victor Ryzhenkin proposed openstack-infra/project-config: Fix for gate-murano-congress-devstack-dsvm https://review.openstack.org/162971 | 18:59 |
anteaya | and holding elections is a check box | 18:59 |
*** craigbr has joined #openstack-infra | 18:59 | |
jroll | sure, so the problem is they need to hold an election, no? | 18:59 |
*** _nadya_ has joined #openstack-infra | 18:59 | |
openstackgerrit | Somay Jain proposed openstack-infra/jenkins-job-builder: Adding more configurable options in Notifications plugin https://review.openstack.org/163137 | 18:59 |
jroll | (also, I think stackforge is just as much a part of openstack as the rest, but that's a different larger perception problem) | 18:59 |
anteaya | jroll: they need to be _seen_ holding an election | 19:00 |
anteaya | anyone can hold an election | 19:00 |
clarkb | meeting ish time (keystone isn't quite done yet) | 19:00 |
jroll | mhmm, and the mailing list isn't enough visibility? | 19:00 |
openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard-webclient: Moved template into correct directory. https://review.openstack.org/163138 | 19:00 |
* jogo elects to have lunch | 19:00 | |
jeblair | jroll: i agree, it is my hope/understanding that the big tent is correcting that problem (stackforge = openstack) | 19:00 |
*** Somay has joined #openstack-infra | 19:00 | |
anteaya | having every stackforge hit the mailing list at the same time the tc ptl elections occur might confuse the electorate | 19:00 |
jeblair | clarkb: and keystone is done | 19:00 |
jroll | jeblair: indeed | 19:01 |
jroll | anteaya: ok, I see now. thanks. | 19:01 |
*** dizquierdo has quit IRC | 19:01 | |
anteaya | jroll: np | 19:01 |
*** markvoelker_ has quit IRC | 19:01 | |
*** emagana has quit IRC | 19:01 | |
openstackgerrit | Merged openstack-infra/project-config: Enable Tempest in Designate dsvm jobs https://review.openstack.org/163118 | 19:02 |
*** melwitt_ has quit IRC | 19:02 | |
*** emagana has joined #openstack-infra | 19:02 | |
*** melwitt has joined #openstack-infra | 19:02 | |
*** ihrachyshka has quit IRC | 19:02 | |
*** craigbr has joined #openstack-infra | 19:03 | |
*** claudiub has quit IRC | 19:04 | |
*** dboik_ has quit IRC | 19:04 | |
*** baoli has joined #openstack-infra | 19:04 | |
*** dboik has joined #openstack-infra | 19:04 | |
*** craigbr has quit IRC | 19:05 | |
*** sarob has joined #openstack-infra | 19:06 | |
*** emagana has quit IRC | 19:06 | |
*** mpopow_ has quit IRC | 19:07 | |
*** jaypipes is now known as jay-afk | 19:09 | |
*** jay-afk is now known as jaypipes-afk | 19:10 | |
tristanC | anteaya: jogo: thanks for the Heads up! To sum-up, for new programs (and for already accepted then), we need to describe on wiki/Election_Officiating_Guidelines a process that scale and that works for everyone | 19:10 |
greghaynes | waynr: Are you PDX based? | 19:10 |
*** yamamoto has joined #openstack-infra | 19:11 | |
*** _nadya_ has quit IRC | 19:12 | |
anteaya | tristanC: well part of the problem is that our page for holding your own election points people at teh mailing list: https://wiki.openstack.org/wiki/Election_Officiating_Guidelines | 19:12 |
anteaya | tristanC: which does'nt scale and which jogo wanted all of stackforge to be using | 19:12 |
anteaya | so yes, updateing taht page with the as yet to be determined scalable workflow is a good way forward | 19:13 |
*** BharatK has quit IRC | 19:13 | |
waynr | greghaynes: yes i am | 19:15 |
greghaynes | waynr: hey! me too | 19:15 |
waynr | wooo! | 19:15 |
*** yamamoto has quit IRC | 19:15 | |
greghaynes | well, thats convenient ;) | 19:16 |
jogo | anteaya: to clarify I want all stackforge projects to *be able* to use the same process if they choose. not saying they need to. but that is just a side point | 19:16 |
anteaya | jogo: understood | 19:17 |
openstackgerrit | Victor Ryzhenkin proposed openstack-infra/project-config: Fix for gate-murano-congress-devstack-dsvm https://review.openstack.org/162971 | 19:17 |
*** craigbr has joined #openstack-infra | 19:17 | |
jhesketh | I have to bail out early sorry guys, I'll be back soon | 19:19 |
*** coolsvap is now known as coolsvap|afk | 19:20 | |
*** BharatK has joined #openstack-infra | 19:20 | |
anteaya | jhesketh: thanks | 19:21 |
BharatK | Can anyone please explain the difference between "gate-<jobname>" and "check-<jobname>"? | 19:21 |
clarkb | BharatK: typically the only difference is where the jobs run, it allows us to run check on different nodes for example | 19:22 |
*** erlon has joined #openstack-infra | 19:22 | |
BharatK | clarkb: thank you, in this patch https://review.openstack.org/#/c/162718/ | 19:23 |
BharatK | clarkb: I found check-tempest-dsvm-full-ceph and gate-tempest-dsvm-full-glusterfs | 19:23 |
waynr | greghaynes: i'd be down for meeting up to talk about zuul github integration | 19:23 |
BharatK | clarkb: will it make any difference? | 19:23 |
*** pblaho_ is now known as pblaho | 19:23 | |
greghaynes | waynr: awesome, im WFH so really any time works for me | 19:23 |
greghaynes | waynr: im also not too far from plabs so that neighborhood is good for me | 19:23 |
clarkb | BharatK: usually you should run the check job in the check queue so that if we do change the nodes then it does what we expect | 19:23 |
BharatK | clarkb: Okay I will send the patch to modify it | 19:24 |
BharatK | clarkb: Thanks | 19:24 |
SpamapS | fungi: regarding a fake-io glance, I'm not 100% sure that wouldn't work. :) | 19:25 |
*** HeOS has joined #openstack-infra | 19:25 | |
SpamapS | fungi: but I'm 100% sure a tmpfs backed glance would be fine. :) | 19:25 |
fungi | SpamapS: yep. just means you need enough virtual memory provisioned to hold everything you upload into glance | 19:25 |
SpamapS | fungi: cirros ftw. :) | 19:26 |
fungi | SpamapS: whereas a fake backend could likely get by on a limited capacity (or just generate data on the fly) | 19:26 |
SpamapS | fungi: since I am using fake virt driver I'll have to do the ready-signalling (if there is such a thing in a test..) locally anyway. :-P | 19:26 |
fungi | true | 19:26 |
SpamapS | fungi: yeah a fake one that just feeds back 1024 * '\0' is fine with me if it doesn't interefere with testing. :) | 19:27 |
SpamapS | I don't want to test glance after all, I just want to test nodepool. :) | 19:27 |
*** unicell has left #openstack-infra | 19:27 | |
*** craigbr has quit IRC | 19:28 | |
fungi | right, probably depends on the degree to which you expect to have any sort of interaction with the resultant image | 19:29 |
fungi | which at most would be limited to ready-script code paths | 19:29 |
waynr | greghaynes: how about tomorrow afternoon? | 19:30 |
greghaynes | waynr: works for me | 19:30 |
*** emagana has joined #openstack-infra | 19:31 | |
*** mpopow_ has joined #openstack-infra | 19:31 | |
*** aysyd has quit IRC | 19:33 | |
openstackgerrit | Fatih Degirmenci proposed openstack-infra/jenkins-job-builder: Pass through git commit to triggered jobs https://review.openstack.org/163149 | 19:34 |
openstackgerrit | Bharat Kumar Kobagana proposed openstack-infra/project-config: Using check job instead of gate job https://review.openstack.org/163150 | 19:35 |
*** sarob has quit IRC | 19:36 | |
*** hodos has quit IRC | 19:36 | |
*** teran has joined #openstack-infra | 19:37 | |
dmsimard | Hi guys. I haven't had to do this in a while.. where do we report bugs that occur in CI so that we can run a recheck ? | 19:37 |
fungi | dmsimard: in the projects which broke | 19:38 |
*** vishy has joined #openstack-infra | 19:38 | |
*** asselin_ has quit IRC | 19:38 | |
dmsimard | fungi: But it's a timeout to a download of a ruby gem | 19:38 |
dmsimard | from jenkins | 19:38 |
dmsimard | aka, I could probably just re-run the merge attempt and it would pass | 19:38 |
dmsimard | But back in my day we'd have to report this and do a recheck with that bug # as reference | 19:39 |
fungi | dmsimard: yeah, so you want a bug that's not a bug in our software but in a project we track. we have an openstack-gate project on launchpad now you can use for stuff like that | 19:39 |
clarkb | also we don't track bugs in the recheck strings any longer | 19:39 |
dmsimard | Right. I thought this had move to storyboard or something, hence why I was lost | 19:39 |
fungi | still makes it possible to link up elastic-recheck to the failure pattern that way | 19:39 |
clarkb | instead the goal is to push all that into elastic rechek as much as possible since it does an overall betterjob tracking the data | 19:39 |
anteaya | clarkb or fungi when you get a chance can you add gothicmindfood to https://review.openstack.org/#/admin/groups/440,members | 19:40 |
anteaya | I'm teaching her the wonders of gerrit | 19:40 |
dmsimard | So what's the flow ? Report a bug, commit something into elastic recheck and then rerun the merge ? | 19:41 |
fungi | anteaya: gothicmindfood: done | 19:41 |
anteaya | fungi: thank you | 19:41 |
clarkb | dmsimard: you can just recheck to start with, then file the bug and add the e-r check | 19:42 |
fungi | dmsimard: http://docs.openstack.org/infra/elastic-recheck/ | 19:42 |
clarkb | dmsimard: you don't need to wait for e-r to have the info before recheking | 19:42 |
*** sabeen has quit IRC | 19:42 | |
dmsimard | ok so the recheck doesn't make a bug # mandatory anymore ? | 19:42 |
*** ociuhandu has quit IRC | 19:43 | |
AJaeger_ | dmsimard: correct | 19:43 |
AJaeger_ | dmsimard: the current state is documented in http://docs.openstack.org/infra/manual/developers.html | 19:43 |
*** fdegir has joined #openstack-infra | 19:44 | |
dmsimard | Ok, thanks for the refresh - haven't seen an actual CI "bug" in a while so I guess that's a good sign :) | 19:44 |
openstackgerrit | Merged openstack-infra/project-config: Fix zuul log path again https://review.openstack.org/163134 | 19:45 |
*** claudiub has joined #openstack-infra | 19:45 | |
*** ominakov_ has joined #openstack-infra | 19:49 | |
*** aysyd has joined #openstack-infra | 19:49 | |
*** otter768 has joined #openstack-infra | 19:51 | |
*** dims has quit IRC | 19:51 | |
*** gokrokve has joined #openstack-infra | 19:51 | |
openstackgerrit | Victor Ryzhenkin proposed openstack-infra/project-config: Fix for gate-murano-congress-devstack-dsvm https://review.openstack.org/162971 | 19:52 |
*** otter768 has quit IRC | 19:55 | |
*** tnovacik has joined #openstack-infra | 20:00 | |
openstackgerrit | Doug Hellmann proposed openstack-infra/project-config: import aeromancer to stackforge https://review.openstack.org/161447 | 20:00 |
openstackgerrit | Doug Hellmann proposed openstack-infra/project-config: Add detail to gerrit ACL normalization errors https://review.openstack.org/161914 | 20:00 |
anteaya | the only part I felt needed to be discussed by infra was whether or not we had foundation db access | 20:00 |
*** dims has joined #openstack-infra | 20:01 | |
anteaya | since that was the only element that might have changed in the electoral roll generation definition | 20:01 |
*** bswartz has quit IRC | 20:01 | |
*** AJaeger_ has quit IRC | 20:01 | |
*** hdd_ has quit IRC | 20:02 | |
fungi | anteaya: the infra root admins do technically have that access, but i don't think we want to use it directly. we need an api | 20:02 |
*** craigbr has joined #openstack-infra | 20:02 | |
clarkb | zaro: https://review.openstack.org/154305 https://review.openstack.org/154300 and https://review.openstack.org/#/c/159159/ | 20:02 |
anteaya | fungi: okay thanks | 20:02 |
anteaya | that was what I needed | 20:02 |
clarkb | fungi: pleia2 jhesketh mordred jeblair SergeyLukjanov ^ that should fix the project creation racing with ansible | 20:02 |
dougwig | infra cores, could we get some reviews on jeblair's idea to bump the experimental priority queue? we're trying to get several jobs done in time for kilo. https://review.openstack.org/#/c/163101/ | 20:02 |
anteaya | what are the chances said api will exist prior to vancouver? | 20:02 |
SergeyLukjanov | clarkb, ack | 20:03 |
clarkb | anteaya: fungi: it mostly already does exist, there is the thing that gerrit checks | 20:03 |
tristanC | Also I proposed a change to implement the zuul_split specs ( https://review.openstack.org/#/c/152290/ ) | 20:04 |
anteaya | does if have the ability to return a list of current foundation individual members that can be relied on? | 20:04 |
clarkb | dougwig: fwiw the old check item is not much newer than the oldest experimental item | 20:04 |
fungi | clarkb: true, though for that i think we also need to narrow scope further. you basically need to know the primary e-mail address of the potential foundation member to test it via that method, and that api actually updates the foundation database when you call it, so not the best current avenue either | 20:05 |
dougwig | clarkb: 7 hours for a client test job, in a queue 7? that's starvation. | 20:05 |
clarkb | fungi: ya its not perfect, but I think 95% of the code is written | 20:05 |
clarkb | dougwig: and 4 hours for check... | 20:05 |
fungi | anteaya: it returns nothing other than an indication of whether the e-mail address you submitted corresponds to a current foundation member or not | 20:05 |
clarkb | dougwig: I am +2 on the chnge | 20:05 |
clarkb | I just don't expect it to solve this problem magically | 20:06 |
dougwig | clarkb: ok, then i'll stop arguing. :) | 20:06 |
clarkb | the underlying issue is that our clouds are not giving us nodes fast enough | 20:06 |
clarkb | if you want to fix the problem fix ^ | 20:06 |
fungi | anteaya: doesn't return a member id number or profile url or anything, it just returns pass or fail | 20:06 |
*** thedodd has quit IRC | 20:06 | |
* clarkb notes that leaky ports are part of the problem there | 20:06 | |
dougwig | clarkb: 4 hours but running tests, versus 7+ and still queued. i'd take 4 hours today. | 20:07 |
dougwig | and tell me who to go beat up at HP to fix the underlying issue. i'll get my bat. | 20:07 |
clarkb | unfortunately we don't have enough info from the cloud to know how or why the ports are not being deleted | 20:07 |
clarkb | dougwig: its more liklely that we need to beat neutron | 20:07 |
clarkb | or nova | 20:07 |
anteaya | fungi: hmmm, that hasn't been able to pass ttx's interpretation in the past | 20:07 |
banix | anyone around who could review this patch? It had two +2s before I had to rebase. Thanks. https://review.openstack.org/#/c/162681/ | 20:07 |
anteaya | fungi: in that foundation members get ballots | 20:07 |
*** SumitNaiksatam has joined #openstack-infra | 20:07 | |
anteaya | fungi: since I do believe we tried that argument last round | 20:08 |
fungi | anteaya: what hasn't been able to pass ttx's interpretation of what in the past? | 20:08 |
anteaya | and i recall him saying to me that the definition wasn't foundation members whose emails match gerrit | 20:08 |
clarkb | lunchtime, back in a bit | 20:10 |
anteaya | yes, I'm still here to continue this but I do also need to eat something | 20:10 |
fungi | anteaya: the current electorate rolls are generated by checking for changes merged to corresponding projects since the specified the validity date and building a list of the gerrit accounts which owned those changes | 20:10 |
openstackgerrit | Bharat Kumar Kobagana proposed openstack-infra/project-config: Using check job instead of gate job https://review.openstack.org/163150 | 20:10 |
*** aysyd has quit IRC | 20:11 | |
fungi | anteaya: the foundation member assumption is made based on the fact that currently those projects are covered by a cla and to sign that cla through gerrit you must have first established a foundation member account with the same e-mail address, though the addresses in both the foundation system and gerrit can diverge thereafter | 20:11 |
anteaya | fungi: yes | 20:11 |
anteaya | and that has worked | 20:11 |
*** unicell has joined #openstack-infra | 20:11 | |
anteaya | yes | 20:12 |
*** ghostpl_ has quit IRC | 20:12 | |
*** thedodd has joined #openstack-infra | 20:12 | |
*** Somay has quit IRC | 20:14 | |
*** carl_baldwin has joined #openstack-infra | 20:14 | |
fungi | anteaya: i don't question whether an api to check foundation membership is useful, but curious what specific use case you're seeing for it in the current tooling | 20:14 |
openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard-webclient: Added sane result set indicators. https://review.openstack.org/162265 | 20:15 |
fungi | for example, i see it as necessary so we can stop requiring foundation membership when agreeing to the icla in gerrit, however we still need something to solve the e-mail address divergence problem if we separate them (for example, migrating gerrit to authenticate to openstackid.org) | 20:15 |
anteaya | if we can reliably select the subset of all reviewers who are _still_ foundation members we may have a smaller electorate and have a greater voter turnout percentage | 20:16 |
anteaya | sorry not all reviewers | 20:16 |
anteaya | all contributors | 20:16 |
fungi | anteaya: i thought last time around we foudn a clause that said that one way you continue to be a foundation member is by contributing code to an official openstack project | 20:17 |
fungi | which made that concern moot | 20:17 |
anteaya | fungi: we did | 20:17 |
anteaya | and last week while you were away jeblair said he disagrees with that | 20:17 |
*** sergsh has joined #openstack-infra | 20:17 | |
anteaya | and I said fine, but for me the only thing we can do if we want to push for a new definition is to poll the foundation db reliably | 20:18 |
fungi | hence we can still (currently) assume that if you have an icla signed and contributed code during the validity period we can continue (for now) to assume they're foundation members for the purpose of holding an election | 20:18 |
*** banix has quit IRC | 20:18 | |
fungi | but yes, if there's another reasonable interpretation of the bylaws which counters that assertion, then we should revisit | 20:19 |
anteaya | fungi: I'm standing on the same hilltop you are | 20:19 |
anteaya | looking in the same direction | 20:19 |
anteaya | as of last week jeblair was not | 20:19 |
openstackgerrit | Max Rydahl Andersen proposed openstack-infra/jenkins-job-builder: Add support for 'raw' to inject direct xml to job configs. https://review.openstack.org/160158 | 20:19 |
anteaya | but he is much better at articulating his point of view than I am | 20:20 |
anteaya | and he is in a meeting and I'm hungry | 20:20 |
anteaya | so I'm willing to take a break for now | 20:20 |
anteaya | and thanks | 20:20 |
*** BobBall is now known as BobBall_AWOL | 20:20 | |
*** banix has joined #openstack-infra | 20:20 | |
fungi | however, i think if that's in question, then it's worth pointing out we have (in my opinion) even bigger gaps between what an actual atc is per the bylaws and what we can reasonably query from our current data | 20:20 |
openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard-webclient: Removed grunt:server https://review.openstack.org/163165 | 20:20 |
*** ociuhandu has joined #openstack-infra | 20:21 | |
fungi | so it's a best-effort approximation regardless. some combination of additional tooling and a clearer redefinition of what an atc is would be needed to converge them | 20:21 |
*** doug-fish has left #openstack-infra | 20:21 | |
*** hdd_ has joined #openstack-infra | 20:22 | |
*** yfried_ is now known as yfried|afk | 20:22 | |
*** aysyd has joined #openstack-infra | 20:22 | |
*** thedodd has quit IRC | 20:23 | |
anteaya | I'm still standing beside you on the same hilltop looking in the same direction | 20:24 |
*** sergsh has quit IRC | 20:24 | |
anteaya | while the oven preheats for my pizza | 20:24 |
tristanC | fungi: anteaya: moreover, jeblair disagreement needs to be present on the charter, because the definition here sounds pretty clear http://git.openstack.org/cgit/openstack/governance/tree/reference/charter.rst#n130 | 20:24 |
anteaya | tristanC: well we have more than just the charter in play, we have the foundation bylaws too | 20:25 |
anteaya | and that other document I found last year | 20:26 |
* anteaya looks it up again | 20:26 | |
gothicmindfood | jogo: thanks for the add! | 20:26 |
anteaya | the technical committe member policy: http://lists.openstack.org/pipermail/openstack-tc/2014-September/000820.html | 20:27 |
anteaya | tristanC: part of the problem is we have different sets of rules | 20:27 |
anteaya | and then interpretations of them | 20:27 |
gothicmindfood | fungi: I mean, thanks for the add :) | 20:28 |
lifeless | headsup - new testtools release about an hour ago. No fallout expected this time (its a very shallow change) | 20:28 |
*** yfried|afk is now known as yfried_ | 20:29 | |
*** kgiusti has left #openstack-infra | 20:29 | |
gothicmindfood | fungi: anteaya mentioned that membership to the group third-party CI list means that one shouldn't get an email notification of testing from those systems after submitting a patch | 20:29 |
tristanC | anteaya: could we repost that mail of your, to make sure that this interpretation still hold for the upcomming elections ? | 20:30 |
gothicmindfood | fungi: but I seem to be getting emails from some of the systems listed there after submitting a test patch to the sandbox repo | 20:30 |
anteaya | tristanC: you can do as you wish as you are an election official this time and I am not, though I would encourage you to consult with your co-election official | 20:32 |
*** ominakov_ has quit IRC | 20:32 | |
anteaya | tristanC: personally I don't see the point in having the same conversation prior to every election unless something has changed | 20:33 |
anteaya | in either tooling or documentation (bylaws or charter) | 20:33 |
anteaya | and it hasn't | 20:33 |
tristanC | agreed, so maybe jeblair can confirm we are on track here ? | 20:34 |
anteaya | well he can offer his opinion yes | 20:34 |
anteaya | but you are the election official | 20:34 |
fungi | gothicmindfood: i'll take a look in a little while. trying to follow the tc meeting but they talk faster than i can read | 20:34 |
anteaya | and need to follow your own insticnts on what is right | 20:34 |
anteaya | to conduct a fair honest equitible election | 20:34 |
*** radez is now known as radez_g0n3 | 20:35 | |
gothicmindfood | fungi: np :) just keep me posted - I'd love to help out/dig in :) | 20:35 |
*** tnovacik has quit IRC | 20:38 | |
*** sarob has joined #openstack-infra | 20:39 | |
*** ominakov_ has joined #openstack-infra | 20:39 | |
*** hdd_ has quit IRC | 20:39 | |
*** emagana has quit IRC | 20:41 | |
*** hdd_ has joined #openstack-infra | 20:41 | |
*** mpopow_ has quit IRC | 20:42 | |
*** boris-42 has quit IRC | 20:42 | |
*** craigbr has quit IRC | 20:43 | |
*** baoli has quit IRC | 20:45 | |
*** dboik_ has joined #openstack-infra | 20:45 | |
*** doug-fish has joined #openstack-infra | 20:46 | |
bknudson | looks like the gate isn't using testtools=1.7.0 yes? | 20:46 |
bknudson | yet? | 20:46 |
*** esker has joined #openstack-infra | 20:46 | |
*** melwitt has quit IRC | 20:46 | |
*** melwitt has joined #openstack-infra | 20:47 | |
bknudson | (wondering if we should add a !=1.7.0) | 20:48 |
*** Somay has joined #openstack-infra | 20:48 | |
fungi | bknudson: we don't seem to have mirrored it http://pypi.dfw.openstack.org/simple/testtools/ | 20:48 |
*** dboik has quit IRC | 20:49 | |
*** koolhead17 has quit IRC | 20:49 | |
*** dprince has quit IRC | 20:49 | |
openstackgerrit | Brant Knudson proposed openstack/requirements: Skip testtools 1.7.0 https://review.openstack.org/163173 | 20:50 |
fungi | bknudson: is it broken somehow? | 20:51 |
bknudson | fungi: I get 7 unit tests fail in keystone. | 20:51 |
bknudson | like this: http://cdn.pasteraw.com/8wc90sw9lh8zdkg3zi8uk13yg1310pj | 20:51 |
bknudson | using tox -e py27 | 20:51 |
fungi | bknudson: interesting. do you know why yet (other than you only see them with newest testtools)? have you opened a bug or mentioned it to lifeless since i think he asked to be kept abreast of failures from that new release? | 20:52 |
bknudson | fungi: not yet. | 20:53 |
*** mpavone_ has quit IRC | 20:55 | |
openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck: Add query for tempest heat slow bug 1430524 https://review.openstack.org/163179 | 20:56 |
openstack | bug 1430524 in tempest "tempest.api.orchestration.stacks.test_neutron_resources.NeutronResourcesTestJSON failures spiking with timeouts since 3/10" [High,In progress] https://launchpad.net/bugs/1430524 - Assigned to Matt Riedemann (mriedem) | 20:56 |
*** banix has quit IRC | 20:57 | |
clarkb | ianw: I am starting to read the nodepool config checker change. Do you think we should try to hammer out what is required and what is optional and so on in this change or do that in a followup? | 20:58 |
clarkb | ianw: I think I prefer to get this in as is then refine simply because the change is large enough as is, but wasn't sure if you had considered this yet | 20:58 |
*** changbl has quit IRC | 20:58 | |
lifeless | bknudson: hi, 'sup ? | 21:00 |
bknudson | lifeless: running unit tests with testtools 1.7.0 is failing several tests... | 21:01 |
bknudson | here's an example: http://cdn.pasteraw.com/8wc90sw9lh8zdkg3zi8uk13yg1310pj | 21:01 |
lifeless | thats odd | 21:01 |
lifeless | its imported _compat3x | 21:01 |
lifeless | on 2.7 | 21:01 |
lifeless | verrry odd | 21:02 |
mriedem | bknudson: do you have a link to a jenkins run with the failure? | 21:02 |
bknudson | mriedem: testtools 1.7.0 isn't in gate yet. | 21:02 |
fungi | lifeless: mriedem: bandersnatch seems to have not updated it yet. i'm about to try and figure out why it's stuck | 21:03 |
lifeless | bknudson: did 1.6.1 have this issue ? | 21:03 |
bknudson | lifeless: no, I switched in my tox env and no errors. | 21:03 |
lifeless | thats bizarre | 21:03 |
lifeless | because - look at the commits | 21:04 |
lifeless | https://github.com/testing-cabal/testtools/commits/master | 21:04 |
lifeless | bknudson: / sigmavirus24: can you do an experiment for me in your environment? | 21:04 |
bknudson | lifeless: yes | 21:05 |
lifeless | start python (the failing version) and then import testtools._compat2x | 21:05 |
lifeless | https://github.com/testing-cabal/testtools/blob/master/testtools/compat.py#L35 is the code that must be going wrong | 21:05 |
bknudson | ImportError: No module named _compat2x | 21:05 |
lifeless | ok so thats bogus | 21:05 |
*** ominakov_ has quit IRC | 21:06 | |
bknudson | import testtools._compat3x -- worked | 21:06 |
lifeless | can you uninstall testtools | 21:06 |
lifeless | and then install testtools with --no-use-wheel | 21:06 |
lifeless | ? | 21:06 |
gordc | ah cool. not just me. | 21:06 |
lifeless | fungi: please don't fix bandersnatch :) | 21:06 |
sigmavirus24 | bknudson: same thing | 21:06 |
lifeless | yeah, but _compat3x is only for 3.x | 21:06 |
bknudson | is it shipped? | 21:06 |
lifeless | it must not be used on 2.x because 2.x doesn't support the things it does | 21:07 |
lifeless | bknudson: they both were in 1.6.1 | 21:07 |
anteaya | I _have_ to go for a walk now, the sun is is just too beautiful | 21:07 |
fungi | lifeless: i'm in a holding pattern (well, mostly paying attention to the cross-project meeting but anyway...) | 21:07 |
lifeless | bknudson: I'm thinking the wheel is damaged | 21:07 |
sigmavirus24 | lifeless: bingo | 21:07 |
sigmavirus24 | lifeless: worked when installed from souroce | 21:07 |
sigmavirus24 | s/souroce/source/ | 21:07 |
*** bhunter71 has quit IRC | 21:07 | |
*** dboik_ has quit IRC | 21:07 | |
bknudson | I only have .tox/py27/lib/python2.7/site-packages/testtools/_compat3x.py | 21:07 |
lifeless | ok I've removed the wheel from pypi | 21:08 |
lifeless | I'll mail the list now | 21:08 |
sigmavirus24 | lifeless: thanks | 21:08 |
lifeless | and file a bug to figure out WTF | 21:08 |
sigmavirus24 | lifeless: side question, how did you generate the wheel? | 21:08 |
lifeless | I built the wheel with python 3.4 vs 2.7 which I normally use | 21:08 |
*** dboik has joined #openstack-infra | 21:08 | |
lifeless | same command line etc etc. | 21:08 |
lifeless | its a universal wheel though | 21:08 |
lifeless | so in principle it doesn't matter. | 21:08 |
lifeless | in practice... | 21:08 |
sigmavirus24 | *shouldn't ;) | 21:09 |
* sigmavirus24 looks at testtools to see what it looks like while glance's tests run | 21:09 | |
*** ihrachyshka has joined #openstack-infra | 21:09 | |
*** sarob has quit IRC | 21:09 | |
*** dkliban is now known as dkliban_afk | 21:10 | |
*** Somay has quit IRC | 21:10 | |
lifeless | mail sent | 21:10 |
*** Somay has joined #openstack-infra | 21:10 | |
lifeless | bknudson: so yeah, no need to skip 1.7.0 :) | 21:11 |
openstackgerrit | Somay Jain proposed openstack-infra/jenkins-job-builder: Adding more configurable options in Notifications plugin -- Support for selecting : * Format * Protocol * Timeout * Log level https://review.openstack.org/163185 | 21:11 |
lifeless | bknudson: also 1.7.0 fixes the triplicate thing, so I think folk will like having it | 21:11 |
bknudson | lifeless: abandoned https://review.openstack.org/#/c/163173/ -- it didn't work anyways. | 21:11 |
bknudson | I think something else was updating testtools | 21:11 |
sigmavirus24 | bknudson: it's coming from within the virtualenv? | 21:12 |
* sigmavirus24 makes bad pop culture reference and feels bad | 21:12 | |
*** ghostpl_ has joined #openstack-infra | 21:13 | |
clarkb | yay for fixing things :) | 21:13 |
*** aysyd has quit IRC | 21:13 | |
*** tnovacik has joined #openstack-infra | 21:14 | |
bknudson | `.tox/py27/bin/pip install testtools` seems to have worked for me now. | 21:14 |
lifeless | yes, cause the wheel is gone :) | 21:14 |
sigmavirus24 | lifeless: I'm not quite sure what https://github.com/testing-cabal/testtools/blob/master/setup.py#L18 is doing | 21:14 |
*** emagana has joined #openstack-infra | 21:14 | |
lifeless | looks like a well intentioned fix syntax error on 3.x patch | 21:15 |
sigmavirus24 | lifeless: looks like _compat2x.py causes problems when installing testtools on py3. So I guess if you're using setup.py bdist_wheel it'll get ignored but that should only happen during installation, not during building of the package | 21:15 |
lifeless | it might be the cause | 21:15 |
lifeless | yah | 21:16 |
sigmavirus24 | It's skipping the module | 21:16 |
openstackgerrit | Clark Boylan proposed openstack-infra/project-config: Add option to set swift delete after header https://review.openstack.org/163188 | 21:16 |
clarkb | jhesketh: ^ something like that maybe? | 21:16 |
jeblair | anteaya, tristanC: it seems very clear to me that, based on the TC charter, you must be both an individual member and have a change merged to be an ATC (or, separately, be confirmed as an extra atc, but let's set that aside for this discussion) | 21:16 |
jeblair | anteaya, tristanC: the tc member policy in the bylaws is poorly worded | 21:17 |
openstackgerrit | Merged openstack-infra/elastic-recheck: Add query for tempest heat slow bug 1430524 https://review.openstack.org/163179 | 21:17 |
openstack | bug 1430524 in tempest "tempest.api.orchestration.stacks.test_neutron_resources.NeutronResourcesTestJSON failures spiking with timeouts since 3/10" [High,In progress] https://launchpad.net/bugs/1430524 - Assigned to Matt Riedemann (mriedem) | 21:17 |
jeblair | anteaya, tristanC: "An Individual Member is an ATC who has had a contribution approved for inclusion..." reads poorly for me | 21:17 |
mriedem | tjat | 21:18 |
mriedem | thanks openstack bot | 21:18 |
openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: Allow any project to make use of the proposal bot script https://review.openstack.org/163189 | 21:18 |
jeblair | anteaya, tristanC: it seems like it is written backwards, and i legitimately do not understand the english construction there. | 21:18 |
clarkb | notmyname: if you have a moment any gotchas with X-Delete-After other than we cannot change the value after we upload the object? | 21:18 |
*** bhuvan has quit IRC | 21:18 | |
lifeless | sigmavirus24: bug 1430534 | 21:18 |
openstack | bug 1430534 in testtools "Wheels built on Python 3.4 are missing _compat2x" [Undecided,New] https://launchpad.net/bugs/1430534 | 21:18 |
*** bhuvan has joined #openstack-infra | 21:19 | |
*** rlandy has quit IRC | 21:19 | |
jeblair | anteaya, tristanC: that sentence should be setting up the requirements for being an atc based on individual membership, but it seems to be setting up requiremnts for individual membership based on being an atc. that's nonsensical because it is in a section called "Determination of Active Technical Contributor." | 21:19 |
*** ldnunes has quit IRC | 21:20 | |
*** thedodd has joined #openstack-infra | 21:20 | |
*** ghostpl_ has quit IRC | 21:20 | |
*** carl_baldwin has quit IRC | 21:20 | |
jeblair | anteaya, tristanC: the only way i read it that makes sense is that it says that in order to be an atc you have to be an individual member with a change included. | 21:20 |
lifeless | sigmavirus24: putting up PR now | 21:20 |
sigmavirus24 | lifeless: yeah I don't recall my last dive into distutils well enough to know if/when build_py is used | 21:21 |
sigmavirus24 | I wonder if there's an option that could be passed to build_module | 21:21 |
* sigmavirus24 loks | 21:21 | |
jeblair | anteaya, tristanC: i suspect it is a typo and should say "An Individual Member is an ATC _if they have_ had a contribution approved". but unfortunately, i can't patch that. | 21:21 |
clarkb | hrm I should use delete_after instead of expire_after for consistency with swift | 21:21 |
jeblair | fungi: ^ sorry, fyi conversation with anteaya and tristanC | 21:22 |
fungi | jeblair: yep | 21:22 |
sigmavirus24 | lifeless: I have an idea | 21:22 |
*** carl_baldwin has joined #openstack-infra | 21:22 | |
*** boris-42 has joined #openstack-infra | 21:22 | |
openstackgerrit | Clark Boylan proposed openstack-infra/project-config: Add option to set swift delete after header https://review.openstack.org/163188 | 21:23 |
sigmavirus24 | lifeless: sending PR shortly | 21:23 |
lifeless | sigmavirus24: backing that out doesn't fix it FWIW | 21:24 |
sigmavirus24 | interesting | 21:24 |
sigmavirus24 | yeah I wasn't sure what build_py was or where it came from | 21:24 |
lifeless | nope I have a fix | 21:25 |
*** Somay has quit IRC | 21:25 | |
lifeless | cached data confusing :) | 21:25 |
*** mpopow has joined #openstack-infra | 21:26 | |
lifeless | sigmavirus24: https://github.com/testing-cabal/testtools/pull/132 | 21:26 |
*** stevemar has quit IRC | 21:27 | |
*** stevemar has joined #openstack-infra | 21:27 | |
jeblair | anteaya, tristanC, fungi: so anteaya's mail interpreted tc memebrship policy 3.(b)(i) with its plain meaning of defining what an individual member is (in terms of an ATC). i contest it can not be doing that because it's actual purpose is to define what an ATC is. | 21:28 |
openstackgerrit | Somay Jain proposed openstack-infra/jenkins-job-builder: Adding more configurable options in Notifications plugin -- Support for selecting : * Format * Protocol * Timeout * Log level https://review.openstack.org/163137 | 21:29 |
*** dannywilson has quit IRC | 21:30 | |
jeblair | fungi, zaro: want to talk about utf8? | 21:30 |
openstackgerrit | Swaminathan Vasudevan proposed openstack-infra/devstack-gate: Configure sub-nodes for Neutron DVR aiopcpu https://review.openstack.org/158525 | 21:30 |
*** pc_m has quit IRC | 21:31 | |
openstackgerrit | Swaminathan Vasudevan proposed openstack-infra/project-config: Create Neutron DVR experimental job for aiopcpu https://review.openstack.org/158415 | 21:31 |
fungi | jeblair: it also, however, even setting aside the probable typo, implies that you can continue being an atc for 365 days after your last contribution regardless of individual membership status (at least it doesn't address the alternative there) | 21:31 |
*** yfried_ is now known as yfried|afk | 21:31 | |
*** dustins has quit IRC | 21:31 | |
fungi | i would argue that the tc member policy entirely lacks any reference to how former individual members who are actively contributing patches to official projects would be treated | 21:32 |
fungi | but also, we lack tooling to actually map them up anyway | 21:33 |
jeblair | fungi: i agree (unless the correction to the probable typo changes that; there's a weird vibe of individual membership infused in that sentence that, since it doesn't make sense to me, i can't work out which parts it's trying to apply to) | 21:33 |
*** andreykurilin_ has joined #openstack-infra | 21:34 | |
fungi | also the lack of specificity around "contribution approved for inclusion" is problematic in my opinion | 21:34 |
fungi | but we should be similarly careful to not get specific about things we can't actually measure/demonstrate either | 21:35 |
*** Somay has joined #openstack-infra | 21:36 | |
jeblair | fungi: the interpretation that makes most sense to me and seems to match the spirit of everything, is that membership is not required for contribution. if you contribute and also choose to become a member, you can vote. if you resign membership, you can not vote. you get to vote for 1 year after your last contribution. | 21:36 |
*** tonytan4ever has quit IRC | 21:36 | |
jeblair | i believe that the existing texts all support that. however, i readily grant they are unclear, and reasonable people may believe they support something different. :) | 21:36 |
*** dizquierdo has joined #openstack-infra | 21:37 | |
tristanC | jeblair: thanks for the explanation. So it means we can't just check gerrit history, we also need to validate membership... | 21:37 |
fungi | i think that's a reasonable and logical interpretation, likely to be the intent of its authors, however one we currently can't use in some corner cases | 21:37 |
*** sdake_ has quit IRC | 21:37 | |
clarkb | ianw: reviewed the nodepool config checker, +2 but with some comments for followup improvements if you are interested | 21:37 |
openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: Add job to keep neutron child repo unit test dependencies up to date https://review.openstack.org/163199 | 21:37 |
tristanC | jeblair: out of curiosity, do you know how big is the difference between contributor without membership versus contributor with membership ? | 21:37 |
fungi | tristanC: or it means we need to be honest that what we're doing is our best approximation of the requirements set forth, but that the requirements themselves can't be fulfilled exactly as stated | 21:37 |
fungi | after all, this is a legal document, and law is mostly about intent and doing the best you can, not followinf a rigid specification | 21:38 |
jeblair | tristanC: i want to be really clear here in that while i believe in my interpretation, i do think it's really unclear, and you should not take my word for it, but rather, we should post some interpretation publicly to a list and get agreement. | 21:39 |
jeblair | ianal | 21:39 |
fungi | as programmers, we're conditioned to think of legal documents as technical specifications which need to precisely match inputs to outputs and actions to reactions | 21:39 |
*** tsg has joined #openstack-infra | 21:39 | |
clarkb | dougwig: I am not sure I understand 163199 we already keep neutron up to date | 21:40 |
*** mattfarina has quit IRC | 21:40 | |
*** sdake_ has joined #openstack-infra | 21:40 | |
*** chlong has joined #openstack-infra | 21:40 | |
fungi | tristanC: not only do we not know how many current code contributors to official projects are no longer foundation individual members, we currently have no clear way to even find out | 21:41 |
clarkb | dougwig: also if this is doing what I think it is doing we should just have those repos as part of global requirements management and not neutron requirements management | 21:41 |
dougwig | clarkb: it's grosser than that. i have to run, but i'll ping you later. what i'm doing may be too gross for words, and i was just exploring it. | 21:42 |
fungi | tristanC: there are some hints in the foundation member database, but due to divergence in the data tracked there and in gerrit there are bound to be members who are active technical contributors but for whom we can't (easily) find their member profile | 21:42 |
clarkb | dougwig: I think we have already solved it with global reqs | 21:42 |
dougwig | clarkb: yes, but we've also decided neutron should not be in global reqs. | 21:43 |
clarkb | dougwig: yes I think this is pointing to underlying decisions that should be reconsidered | 21:43 |
clarkb | dougwig: I really don't want to reengineer the same problem twice to accomodate a one off | 21:43 |
zaro | jeblair: sure, ready when you are | 21:44 |
dougwig | clarkb: absolutely true. let me get back with you after my appt. | 21:44 |
clarkb | dougwig: kk | 21:44 |
fungi | tristanC: it's also worth noting that our current process is analogous to a government putting together a list of all eligible voters based on census data, registering them and sending ballots to their homes | 21:44 |
*** mriedem has quit IRC | 21:44 | |
clarkb | fungi: like australia except we don't fine you if you don't vote | 21:45 |
fungi | tristanC: when many governments expect you to register your intent to vote and go to a voting place, identify yourself and then cast your ballot | 21:45 |
jeblair | fungi, zaro: okay, here's my vague discomfort around using the connection string rather than table metadata: | 21:46 |
jeblair | fungi, zaro: if we do that, we're keeping information about how to interpret what is in the db outside of mysql | 21:46 |
jeblair | maybe that's okay to do, but if we can do the same thing by setting the table metadata correctly, shouldn't we consider that first? | 21:47 |
fungi | jeblair: i apparently misread the earlier discussion because i thought zaro said he tried changing the table metadata and it made no difference to gerrit's behavior, but that changing the dburi (at pelix's recommendation) had the desired effect | 21:47 |
*** tiswanso has quit IRC | 21:47 | |
jeblair | fungi, zaro: so is the effect of changing the table metadata known or unknown? | 21:48 |
fungi | now unknown to me, since in the meeting i think zaro said he hadn't tried it after all | 21:48 |
zaro | that is correct. i did not try at all | 21:49 |
asselin | just started hitting this error? ring any bells? haven't investigated yet: http://paste.openstack.org/show/191416/ | 21:49 |
*** ominakov_ has joined #openstack-infra | 21:49 | |
jeblair | zaro: can you try removing the dburi and changing the table metadata on review-dev? that way at least we have full info | 21:49 |
zaro | ok. will give that a try as well. | 21:49 |
tristanC | fungi: jeblair: got it, so considering: the foundation db is not consistent with gerrit's + the different possible interpretation of bylaws + the fact gerrit db is more usable and represent better actual contributor, the best way out of it is to proceed like previous elections | 21:49 |
ianw | clarkb: sorry, re previous comment, you fatally oopsed my kernel :) | 21:50 |
asselin | pkg_resources.VersionConflict: (unittest2 0.5.1 (/usr/lib/python2.7/dist-packages), Requirement.parse('unittest2>=1.0.0')) | 21:50 |
clarkb | ianw: you should get a new kernel :) | 21:50 |
clarkb | ianw: also really? my message made kernel have a sad? | 21:51 |
*** mwagner_lap has quit IRC | 21:51 | |
clarkb | asselin: I think that means you have unittest2 installed globall but it is too old | 21:51 |
fungi | tristanC: that's not to say that we shouldn't try to improve it. but it would help to start out with an ml discussion about our current mechanisms by which we assemble the electorate, how closely it matches a reasonable interpretation of the requirements, where our known gaps are, and how we can make it better in future cycles | 21:51 |
*** yfried|afk is now known as yfried_ | 21:52 | |
ianw | clarkb: i do now ... i had just started a meeting with video/usb headset and switched back to my irc vm and boom. so i think something to do with input layer. but anyway | 21:52 |
*** otter768 has joined #openstack-infra | 21:52 | |
asselin | clarkb, these are on my clean nodepool VMs....I guess I need a new image? aha...seems there's a new one build 0.19 hours ago | 21:52 |
jeblair | fungi, zaro: i also have a specific concern with the values in the dburi -- zaro set the collation to "utf8_unicode_ci", but i believe in the meeting last year we decided on utf8_bin. probably regardless of how we do this (metadata or uri), we should use that collation. | 21:53 |
*** fdegir has quit IRC | 21:53 | |
ianw | clarkb: my preference would be to build up larger logic in the config checker separately and commit them with tests. i can make a few tweaks like the float stuff you mentioned | 21:53 |
tristanC | fungi: agreed, else it will be the same problem for next election... | 21:53 |
jeblair | fungi, zaro: http://eavesdrop.openstack.org/meetings/infra/2014/infra.2014-03-04-19.01.log.html if you want a refresher | 21:53 |
ianw | clarkb: honestly just the yaml parse probably gets 90% of the problems i've caused | 21:53 |
jroll | ianw: seeing you here reminded me, there's a new fedora image on rackspace, should have your fixes | 21:53 |
clarkb | ianw: ya its a major improvement as is | 21:53 |
tristanC | fungi: jeblair: and many thanks again for your valuable inputs, legal science is not exactly my cup of tea :) | 21:53 |
fungi | tristanC: for known corner cases it also helps to ask what the actual harm is. for example will someone be able to game the election by voting with an account which has recent technical contributions but resigned the corresponding individual member account | 21:53 |
clarkb | ianw: so I was happy to +2 it without any changes | 21:53 |
ianw | jroll: cool, i'll check it out. f21 job is still unstable, when we tried to switch to d-i-b versions the world exploded, hence detours into nodepool config-checking etc :) | 21:54 |
jroll | ianw: heh, fun | 21:54 |
*** dkranz has quit IRC | 21:55 | |
jeblair | fungi, tristanC: yeah, honestly, i think we should just do what we've been doing for this election because there should be no real harm even if we're slightly off, and if we've worked out the membership/openstackid/CLA/DCO stuff by next time, hopefully we can get things clarified by then. | 21:55 |
jeblair | or worked out at least some of those things ^ | 21:55 |
openstackgerrit | Spencer Krum proposed openstack-infra/puppet-logstash: Flesh out the readme https://review.openstack.org/153820 | 21:56 |
dansmith | clarkb: can I recheck this? https://review.openstack.org/#/c/162218/ | 21:56 |
zaro | jeblair: i forgot about that. i'll verify that it still works when i change the db url to utf8_bin before doing the db change | 21:56 |
jeblair | zaro: good idea | 21:56 |
clarkb | dansmith: yes I think so | 21:56 |
*** otter768 has quit IRC | 21:56 | |
dansmith | clarkb: okay, the failure is unrelated, but being d-g, I figure I'd ask first | 21:56 |
clarkb | dansmith: looks like it failed on a heat test | 21:56 |
clarkb | dansmith: I thought we had those disabled by default so might be worth investigating, but recheck should be fine | 21:57 |
dansmith | clarkb: oh, it did, I saw neutron and kept on moving :) | 21:57 |
dansmith | but it's heat orchestration of neutron I guess | 21:57 |
fungi | lifeless: i should be safe to troubleshoot/fix bandersnatch on our mirrors now, right? deleting the broken whl file from pypi seems to be all that was needed? | 21:58 |
dansmith | clarkb: either way, nobody is using this library yet, so ... can't be related to that kinda failure unless I'm really really good :) | 21:58 |
clarkb | dansmith: ha | 21:58 |
*** bhunter71 has joined #openstack-infra | 21:59 | |
*** jaypipes-afk is now known as JAYPIPES | 21:59 | |
*** JAYPIPES is now known as jaypipes | 21:59 | |
*** sigmavirus24 is now known as sigmavirus24_awa | 22:00 | |
lifeless | fungi: yup | 22:00 |
mmedvede | asselin: I am seeing the same error with requirements | 22:01 |
mmedvede | asselin: and it also shows here, for example: http://logs.openstack.org/33/161833/9/check/check-grenade-dsvm/01379b2//logs/grenade.sh.txt.gz | 22:01 |
*** yfried_ is now known as yfried|afk | 22:02 | |
*** craigbr has joined #openstack-infra | 22:02 | |
asselin | mmedvede, ok I'm testing my 'new' images now to see if that fixes it. | 22:02 |
clarkb | mmedvede: asselin I think that means python-unittest2 should be removed from devstack's dep list | 22:02 |
openstackgerrit | Diego proposed openstack-infra/infra-manual: Add details on using git-review over https https://review.openstack.org/161078 | 22:03 |
zaro | jeblair: db url with utf8_bin worked as well so i'll start the db conversion. | 22:03 |
*** coolsvap|afk has quit IRC | 22:03 | |
*** tsg has quit IRC | 22:03 | |
tristanC | jeblair: fungi: agreed, it's surely a pendantic discussion, but I guess it's what make an election process better :) | 22:04 |
*** bknudson has quit IRC | 22:04 | |
*** dizquierdo has quit IRC | 22:04 | |
fungi | sometimes it's fun to be a pedant | 22:05 |
openstackgerrit | Diego proposed openstack-infra/infra-manual: Add details on using git-review over https https://review.openstack.org/161078 | 22:05 |
*** ominakov_ has quit IRC | 22:06 | |
*** bodepd has quit IRC | 22:06 | |
*** bhuvan has quit IRC | 22:06 | |
*** nelsnelson has quit IRC | 22:06 | |
*** niska has quit IRC | 22:06 | |
*** ominakov_ has joined #openstack-infra | 22:06 | |
*** ominakov_ has quit IRC | 22:07 | |
*** chlong has quit IRC | 22:07 | |
*** agentle_ has quit IRC | 22:07 | |
asselin | clarkb, mmedvede these here? http://paste.openstack.org/show/191418/ | 22:08 |
*** bodepd has joined #openstack-infra | 22:08 | |
fungi | gothicmindfood: do you have an example of an account in the third-party ci group you've received a gerrit notification on? | 22:08 |
*** adalbas has quit IRC | 22:08 | |
clarkb | asselin: ya but there should be a deb one too since the grenade job runs on ubuntu | 22:08 |
gothicmindfood | fungi: IBM xCAT CI is one | 22:08 |
*** niska has joined #openstack-infra | 22:08 | |
*** bhuvan has joined #openstack-infra | 22:09 | |
gothicmindfood | fungi: Mellanox CI is another | 22:09 |
fungi | gothicmindfood: can you point me to the change corresponding to the notification, if you still have it? | 22:09 |
fungi | i just want to make sure that the account they're commenting with is the same account in that group, rather than merely one with the same display name | 22:10 |
gothicmindfood | fungi: https://review.openstack.org/#/c/163147/ | 22:10 |
asselin | clarkb, mmedvede fyi, tests on my new image (.47hr old) just failed with same error. | 22:10 |
*** andreykurilin_ has quit IRC | 22:10 | |
mmedvede | asselin: yes, all the upcoming checks are going to fail upstream as well | 22:11 |
fungi | gothicmindfood: thanks, see if i can tell what's going on there | 22:11 |
BobBall_AWOL | asselin: Could you add XenProject_CI in the third party mail filter list? | 22:11 |
asselin | BobBall_AWOL, sure | 22:11 |
BobBall_AWOL | Correction - XenProject-CI (not _) | 22:11 |
BobBall_AWOL | Thanks! | 22:12 |
asselin | BobBall_AWOL, can't seem to find that account... | 22:12 |
*** _nadya_ has joined #openstack-infra | 22:12 | |
BobBall_AWOL | Even with the correction? | 22:13 |
asselin | BobBall_AWOL, nothing much starts with 'x'. | 22:13 |
gothicmindfood | fungi: kk. I assume you mean besides me trolling emacs users :) | 22:13 |
BobBall_AWOL | email is openstack-ci@xenproject.org - does that find it? | 22:14 |
*** tsufiev_ has quit IRC | 22:14 | |
*** dteselkin has quit IRC | 22:14 | |
*** rakhmerov has quit IRC | 22:14 | |
*** katyafervent has quit IRC | 22:14 | |
*** amitgandhinz has quit IRC | 22:14 | |
asselin | BobBall_AWOL, nope | 22:14 |
fungi | hrm, well bknudson and mridem have vanished, but anyway it looks like bandersnatch automatically unstuck after the testtools whl was deleted. (21:45 utc run today) strange... | 22:15 |
zigo_ | fungi: sdague: If you didn't notice, I removed the python-nova dependency from novnc, and it migrated to Jessie this way. | 22:15 |
*** dimtruck is now known as zz_dimtruck | 22:15 | |
zigo_ | (as per sdague request) | 22:15 |
*** katyafervent has joined #openstack-infra | 22:15 | |
fungi | zigo_: cool. i don't think i have any history on the problem there. why was novnc depending on python-nova in the first place? | 22:16 |
*** katyafervent has quit IRC | 22:16 | |
BobBall_AWOL | asselin: Account ID is 15286 | 22:16 |
zigo_ | fungi: It used to need it, AFAIR. | 22:16 |
*** _nadya_ has quit IRC | 22:17 | |
*** amitgandhinz has joined #openstack-infra | 22:17 | |
*** katyafervent has joined #openstack-infra | 22:17 | |
fungi | zigo_: ahh, got it. i don't currently manage an openstack deployment (aside from devstack on test nodes) so never really got familiar with how novnc is implemented | 22:17 |
fungi | i had always assumed it was a separate vnc replacement | 22:18 |
zigo_ | novnc stands for nova vnc ... :P | 22:18 |
clarkb | jeblair: the query you posted in today's meeting crashes gerrty with AttributeError: 'NoneType' object has no attribute 'get_children' | 22:18 |
clarkb | jeblair: is that reproduceable by you? | 22:18 |
*** amitgandhinz has quit IRC | 22:19 | |
*** spzala has quit IRC | 22:19 | |
BobBall_AWOL | I always thought the novnc was the HTML5 vnc client that meant you didn't need a local client installed https://kanaka.github.io/noVNC/ | 22:19 |
*** Somay has quit IRC | 22:20 | |
asselin | BobBall_AWOL, I can't search by account id....can you find yourself here and send me the link? https://review.openstack.org/ | 22:20 |
*** hdd_ has quit IRC | 22:20 | |
fungi | zigo_: interesting. i can't find any reference to it sharing a history with openstack and thought it was just something which was stuffed into common deployments as a web-based vnc client choice | 22:20 |
*** rakhmerov has joined #openstack-infra | 22:20 | |
*** hdd_ has joined #openstack-infra | 22:20 | |
BobBall_AWOL | asselin: https://review.openstack.org/#/q/owner:openstack-ci%2540xenproject.org+status:open,n,z ? | 22:21 |
fungi | zigo_: i know for a while we maintained a fork of it because its upstream had vaporized, but once they resurfaced we got rid of our fork | 22:21 |
*** rakhmerov has quit IRC | 22:22 | |
*** katyafervent has quit IRC | 22:22 | |
asselin | BobBall_AWOL, nothing found with that query.... | 22:22 |
BobBall_AWOL | asselin: Well it hasn't submitted any review requests, but has commented, e.g. on https://review.openstack.org/#/c/156793/ | 22:22 |
*** dteselkin has joined #openstack-infra | 22:22 | |
*** bswartz has joined #openstack-infra | 22:23 | |
asselin | BobBall_AWOL, ok found it | 22:23 |
asselin | and done | 22:23 |
*** dteselkin has quit IRC | 22:23 | |
BobBall_AWOL | Perfect - thanks! | 22:23 |
asselin | BobBall_AWOL, https://review.openstack.org/#/admin/groups/270,members | 22:24 |
BobBall_AWOL | yay :) | 22:24 |
*** trey has quit IRC | 22:24 | |
*** agentle_ has joined #openstack-infra | 22:25 | |
asselin | BobBall_AWOL, it was there...I just needed to type a few characters and ignore gerrit's un?helpful completion suggestions. | 22:26 |
*** trey has joined #openstack-infra | 22:26 | |
BobBall_AWOL | I think that sometimes gerrit is trying to be a friend-making service. Suggesting everyone other than the person you're trying to find as a completion. | 22:27 |
clarkb | BobBall_AWOL: ha | 22:27 |
fungi | BobBall_AWOL: we're all friends here | 22:27 |
BobBall_AWOL | :) | 22:27 |
asselin | :) | 22:28 |
*** hodos|2 has joined #openstack-infra | 22:28 | |
jeblair | clarkb: i have a change to fix it, let me merge it | 22:29 |
jeblair | clarkb: or you can lowercase the booleans | 22:29 |
openstackgerrit | Merged stackforge/gertty: Fix searching with uppercase booleans https://review.openstack.org/162361 | 22:29 |
clarkb | jeblair: oh that'll work for now. thank you | 22:29 |
*** rakhmerov has joined #openstack-infra | 22:29 | |
jeblair | clarkb: either way ^ :) | 22:29 |
clarkb | or I can just update :) | 22:29 |
jeblair | clarkb: but if you're upgrading anyway, please try out 163085 | 22:30 |
jeblair | fungi: ^ you too if it's not too much of a bother | 22:30 |
jeblair | that's a good change to have wide feedback on | 22:30 |
*** claudiub has quit IRC | 22:30 | |
fungi | jeblair: you bet | 22:30 |
* clarkb grabs that change | 22:31 | |
*** dteselkin has joined #openstack-infra | 22:32 | |
fungi | gothicmindfood: do you happen to know whether those accounts were recent additions to that group? i wonder if this is yet something else which hinges on gerrit cache data and if a cache flush or eventual expiration solves it | 22:33 |
*** sdake__ has joined #openstack-infra | 22:33 | |
*** katyafervent has joined #openstack-infra | 22:33 | |
*** thedodd has quit IRC | 22:34 | |
openstackgerrit | Merged openstack-infra/storyboard-webclient: Read user preferences from backend https://review.openstack.org/137798 | 22:34 |
clarkb | jeblair: the search thing is fixed, thank yo | 22:34 |
*** hashar has quit IRC | 22:34 | |
*** hashar has joined #openstack-infra | 22:35 | |
*** hashar has quit IRC | 22:35 | |
*** tsufiev_ has joined #openstack-infra | 22:36 | |
fungi | clarkb: looking at the git-fe servers, it appears those were built after the uid/gid remapping change merged but before my login.defs change merged, which explains the problem we have there | 22:36 |
*** bhunter71 has quit IRC | 22:37 | |
fungi | clarkb: so they added the admin and sudoers groups at the next available gids | 22:37 |
*** sdake_ has quit IRC | 22:37 | |
clarkb | fungi: I see | 22:38 |
fungi | clarkb: i'll spin up a dupe of one of those to see if it's better with the login.defs change (my review dupe was fine but that was ubuntu and this is centos) | 22:38 |
fungi | but so far all indications are that my patch did solve the problem and we just have some possible cleanup on servers built in that timeframe | 22:38 |
cinerama | do we currently use openstackid.org? just thinking that most of the services i login to regularly use u1 | 22:38 |
fungi | cinerama: it's recent, and so far only used by groups.openstack.org | 22:39 |
fungi | cinerama: the goal is to move all our other systems to use that instead of ubuntuone/launchpad | 22:39 |
clarkb | fungi: ok, I was probably just very confused then | 22:40 |
*** timcline_ has joined #openstack-infra | 22:40 | |
fungi | clarkb: it was a confusing problem and took me a while to figure out what was going on when i first ran into it | 22:40 |
*** timcline_ has quit IRC | 22:41 | |
jeblair | yeah, i'm on the fence about whether we should use openstackid for zanata. i mostly want to keep our services using launchpad/u1 and switch them all at the same time. however, currently translators are not using it, so we can avoid a double change this way. | 22:41 |
jeblair | it's probably a good idea to go aheand and add another small-ish service to help work out any issues, so i'm in favor of openstackid for it. | 22:42 |
*** timcline has quit IRC | 22:42 | |
fungi | cinerama: i believe the next site slated to start using openstackid.org is the www.openstack.org site (which should be a null migration since they share a backend data source) | 22:42 |
jeblair | but that's an exception, not the rule, i think. :) | 22:42 |
openstackgerrit | Khai Do proposed openstack-infra/puppet-gerrit: Fix Gerrit character encoding https://review.openstack.org/163104 | 22:43 |
cinerama | so my take on it thus far is that i'm having some issues getting it going, but i only just started this afternoon | 22:43 |
jeblair | cinerama: how so? | 22:43 |
fungi | jeblair: agreed, the groups production switch to openstackid pointed out that a recent silverstripe upgrade had started using a different default password hashing algorithm and openstackid didn't know what to do with it, so there are definitely still probably some rough edges | 22:43 |
jeblair | fungi: a definite probability indeed! | 22:44 |
fungi | an almost certain possibility! | 22:44 |
cinerama | 2015-03-10T22:28:02,045Z ERROR [org.openid4java.consumer.ConsumerManager] (default task-2) Association attempt, but no discovery endpoints provided. | 22:44 |
jeblair | is anything using openid with it? | 22:44 |
fungi | jeblair: groups.o.o _was_ until recentlyish when it switched to oauth2 | 22:45 |
*** asettle has quit IRC | 22:45 | |
openstackgerrit | Merged openstack-infra/system-config: Start testing review.o.o on trusty https://review.openstack.org/153063 | 22:45 |
jeblair | ah right, hopefully it hasn't broken in the interim | 22:45 |
fungi | but it was working fine with openid prior to that anyway | 22:45 |
*** asettle has joined #openstack-infra | 22:45 | |
fungi | change volume on the openstackid codebase has been really low since then | 22:45 |
zaro | jeblair: tested some more and noticed that this is the min required setting for utf8 to work: https://review.openstack.org/#/c/163104/2 | 22:45 |
zaro | will also do the db conversion later, but i gotta take off for today. | 22:45 |
openstackgerrit | Steve Leon proposed openstack-infra/project-config: New project: cue-dashboard to stackforge https://review.openstack.org/162804 | 22:45 |
*** ayoung has quit IRC | 22:46 | |
*** agentle_ has quit IRC | 22:47 | |
fungi | jeblair: confirmed gertty release crashes on your priority efforts query but tip of master does now | 22:49 |
fungi | er, tip of master does not | 22:49 |
openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard: Removed Process Pool execution. https://review.openstack.org/161374 | 22:49 |
jeblair | fungi: can you test the webbrower change at 163085? | 22:50 |
fungi | funny how much difference in meaning is conveyed by a one letter difference | 22:50 |
*** asettle has quit IRC | 22:50 | |
clarkb | jeblair: so far I have opened things just fine but I use chromium as browser on this machine | 22:50 |
*** asettle has joined #openstack-infra | 22:50 | |
fungi | ooh, that will be interesting to test | 22:51 |
jeblair | clarkb: cool, knowing it didn't break chromium is good too :) | 22:51 |
fungi | jeblair: seems to open into my graphical browser (iceweasel) on my workstation, presumably courtesy of xdg-open. no stderr spam | 22:52 |
*** dboik_ has joined #openstack-infra | 22:52 | |
fungi | i'll have to fiddle to figure out how to get it to fire up a console browser instead. i'd not previously tried following hyperlinks in reviews | 22:53 |
jeblair | fungi: i have not tried that -- there are two things that may complicate that -- if you use xdg-open, my patch will break it by redirecting std*. but also, i think gertty might still try to write to the screen. then again, maybe the call to open the webbrowser would be syncronous, and it would work. | 22:55 |
jeblair | fungi: https://hg.python.org/cpython/file/2.7/Lib/webbrowser.py will be helpful | 22:55 |
*** eharney has quit IRC | 22:55 | |
*** dboik has quit IRC | 22:55 | |
*** mrmartin has quit IRC | 22:55 | |
fungi | jeblair: oh, fair point | 22:56 |
*** hodos|2 has quit IRC | 22:57 | |
fungi | yeah, my hope would be that the browser process starts and grabs that terminal, and then gertty returns when i exit the browser | 22:57 |
fungi | but chances are that is not what would actually happen | 22:57 |
clarkb | nibalizer: I am trying to dig into the template base merger change now. didn't jesusaurus mention a thing about the slave classes? does that come into play here? | 22:57 |
*** dboik_ has quit IRC | 22:57 | |
jeblair | fungi: BROWSER=links ~/bin/gertty openstack | 22:57 |
jeblair | fungi: that works for me as expected with my patch! :) | 22:58 |
nibalizer | clarkb: so i think the bare nodes use ::template directly | 22:58 |
nibalizer | and so we'd have to refactor that when template rolls into server | 22:58 |
jeblair | fungi: it looks like it knows that links needs the terminal and does indeed block | 22:58 |
nibalizer | but for that change it doesn't effect anything | 22:59 |
nibalizer | im interested to see if people think that is the right direction | 22:59 |
*** claudiub has joined #openstack-infra | 22:59 | |
jeblair | cinerama: i'm wondering if openstackid.org is missing something related to openid discovery | 22:59 |
jeblair | cinerama: you might just try using "https://login.launchpad.net/+openid" to validate things work as expected with zanata | 22:59 |
jeblair | since we know that works as an openid SSO url | 22:59 |
cinerama | jeblair: i'm actually pointed at openstackid-dev.openstack.org as i had issues with certs from openstackid.org | 23:00 |
cinerama | jeblair: yeah, that was what i was thinking i'd do after i'd had a quick look | 23:00 |
openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Update dib script usage https://review.openstack.org/163217 | 23:01 |
clarkb | nibalizer: I see, single_use_slave uses template directly whereas slave.pp (which is for long lived ndoes) uses server.ppy | 23:02 |
*** ihrachyshka has quit IRC | 23:02 | |
clarkb | nibalizer: which make sense considering their roles in testing | 23:02 |
clarkb | nibalizer: why did you not delete base.pp? | 23:03 |
nibalizer | clarkb: :shrug: to prove its empty i guess? can nuke it on the next changeset | 23:03 |
clarkb | nibalizer: I almost think we should nuke it in this one to make sure that things don't break due to inadverdently needing it | 23:04 |
clarkb | nibalizer: the tests won't protect you if that include works | 23:04 |
nibalizer | yea, i agree | 23:04 |
nibalizer | kk ill nuke the base.pp | 23:04 |
nibalizer | any other feedback before i rev? or is it in the change? | 23:05 |
nibalizer | what do you think of the overall idea? | 23:05 |
*** ashleighfarnham has quit IRC | 23:05 | |
jeblair | dhellmann, mtreinish, dtroyer: now that i understand more about devstack-plugins and that devstack-plugin repos are the exception, not the rule, i think i will abandon the devstack-plugin-cookiecutter effort | 23:05 |
cinerama | jeblair: so i get a lot further with LP's SSO | 23:06 |
clarkb | nibalizer: overall I think I am fine with a single "template" thing | 23:06 |
clarkb | simplifies the puppet. I have not gone through the code moves iwth a comb though | 23:07 |
*** melwitt has quit IRC | 23:07 | |
jeblair | cinerama: good; though that is suggesting openstackid.org is missing something related to openid which we will eventually need to figure out | 23:08 |
*** esker has quit IRC | 23:08 | |
*** esker has joined #openstack-infra | 23:09 | |
*** Sukhdev has joined #openstack-infra | 23:10 | |
*** erlon has quit IRC | 23:11 | |
*** sdake__ has quit IRC | 23:11 | |
*** prad has quit IRC | 23:14 | |
*** BharatK has quit IRC | 23:14 | |
*** tnovacik has quit IRC | 23:15 | |
*** tsufiev_ has quit IRC | 23:16 | |
*** dteselkin has quit IRC | 23:16 | |
*** katyafervent has quit IRC | 23:16 | |
*** rakhmerov has quit IRC | 23:16 | |
*** katyafervent has joined #openstack-infra | 23:17 | |
gothicmindfood | fungi: I have no knowledge about recent account adds to that group | 23:20 |
fungi | jeblair: yep, awesome. that does work as i had hoped, with the exception that i need to ctrl-l after terminating the browser | 23:20 |
gothicmindfood | fungi: is there a good way to find that out that I can help out with? | 23:20 |
*** tsufiev_ has joined #openstack-infra | 23:21 | |
fungi | cinerama: what issues did you have with the openstackid.org cert? | 23:21 |
*** hdd_ has quit IRC | 23:22 | |
fungi | gothicmindfood: no, i was mostly hoping that you had added them or something and that's what led you to be concerned it wasn't working on them. the auditing in gerrit is not easy to get at for that sort of stuff | 23:22 |
cinerama | fungi: if you do openssl s_client -connect openstackid.org:443 | 23:22 |
cinerama | "unable to get local issuer" | 23:23 |
*** rakhmerov has joined #openstack-infra | 23:24 | |
*** craigbr has quit IRC | 23:24 | |
*** dteselkin has joined #openstack-infra | 23:24 | |
fungi | cinerama: interesting. do you get the same for, e.g., review.openstack.org? | 23:24 |
*** dangers is now known as dangers_away | 23:25 | |
fungi | i'm still fiddling with the ca options to openssl because i never can remember how to test chain validation without spending a couple minutes poking at it | 23:25 |
*** bdpayne_ has joined #openstack-infra | 23:26 | |
*** camunoz has quit IRC | 23:26 | |
cinerama | fungi: you also need -CApath /etc/ssl/certs/ (or wherever) | 23:27 |
*** ddieterly has quit IRC | 23:27 | |
fungi | yeah, i was still getting an error with that and am trying to track down the correct ca | 23:27 |
*** ddieterly has joined #openstack-infra | 23:27 | |
*** rakhmerov has quit IRC | 23:27 | |
*** tsufiev_ has quit IRC | 23:27 | |
*** katyafervent has quit IRC | 23:27 | |
*** dteselkin has quit IRC | 23:27 | |
cinerama | fungi: so in my browser all is well. with the openssl client openstackid-dev.o.o is fine but openstackid.o is not | 23:27 |
*** katyafervent has joined #openstack-infra | 23:28 | |
*** wenlock has quit IRC | 23:28 | |
fungi | how strange | 23:28 |
*** bdpayne_ has quit IRC | 23:28 | |
cinerama | fungi: i picked it up because wildfly had a sad when i tried to connect to openstackid.org for openid. i suspect the cert chain might not be set up correctly but the certs are in heira | 23:29 |
*** katyafervent has quit IRC | 23:29 | |
openstackgerrit | Joshua Harlow proposed openstack/requirements: Bump up debtcollector to 0.3 https://review.openstack.org/163225 | 23:29 |
dstufft | the chain isn't setup correcly for openstackid.org | 23:29 |
*** bdpayne has quit IRC | 23:29 | |
dstufft | it's not serving any intermediate certs | 23:29 |
mgagne_PHL | I'm getting "SSL certificate problem: unable to get local issuer certificate" | 23:29 |
dstufft | browsers are cheaters | 23:29 |
cinerama | dstufft: yep | 23:30 |
dstufft | they'll look at the AIA information and download missing interemediate certificates | 23:30 |
dstufft | https://www.ssllabs.com/ssltest/analyze.html?d=openstackid.org&latest | 23:30 |
dstufft | ssllabs.com is way easier then mucking with s_client | 23:30 |
dstufft | :D | 23:30 |
dstufft | also those ciphers make me sad | 23:31 |
dstufft | RC4 and no PFS? :( | 23:31 |
fungi | dstufft: yep, i'm poking on the server now to see what's up | 23:31 |
*** cody-somerville has quit IRC | 23:31 | |
fungi | with the chain that is | 23:31 |
fungi | dstufft: patches welcome ;) | 23:31 |
dstufft | well some PFS | 23:31 |
dstufft | not DHE not ECDHE | 23:32 |
*** ddieterly has quit IRC | 23:32 | |
dstufft | fungi: where's it controlled | 23:32 |
dstufft | do I gotta learn puppet | 23:32 |
fungi | dstufft: also running on ubuntu 12.04 so some of your favorite ciphers and hashes may not be implemented | 23:32 |
greghaynes | clarkb: so any protips on what a nodepool upload failure looks like? Im guessing the image from glanceclient.images.create throws a traceback on image.update? | 23:32 |
fungi | dstufft: i'll dig up the patch we did for heartbleed | 23:33 |
*** rakhmerov has joined #openstack-infra | 23:33 | |
*** miqui has joined #openstack-infra | 23:33 | |
clarkb | greghayes ya, I am not actually sure what exception it throws though and its probably different based on glance api version | 23:33 |
clarkb | er greghaynes ^ | 23:33 |
*** david-lyle is now known as david-lyle_afk | 23:33 | |
*** mwagner_lap has joined #openstack-infra | 23:34 | |
greghaynes | yep, but I should be able to RuntimeException and its still a validish test? | 23:34 |
*** rakhmerov has quit IRC | 23:34 | |
clarkb | greghaynes: ya, nodepool is usually pretty broad in its catching of stuff so Exception() would probably eve nwork | 23:34 |
*** garyh has quit IRC | 23:35 | |
clarkb | also I am looking at a local coverage report to see what sorts of things need better testing in nodepool. We should definitely get my cmd tests in as they are completely untested | 23:35 |
*** craigbr has joined #openstack-infra | 23:36 | |
*** garyh has joined #openstack-infra | 23:36 | |
clarkb | gearman and zmq stuff in nodepool.py need some work too as well as some of the jenkins stuff | 23:37 |
fungi | dstufft: a change along the lines of https://review.openstack.org/#/c/128995/1/modules/openstackid/templates/vhost.erb (though that file has now moved into the openstack-infra/puppet-openstackid project in the templates top-level directory) | 23:37 |
*** craigbr has quit IRC | 23:37 | |
*** dteselkin has joined #openstack-infra | 23:37 | |
dstufft | oh is this erb | 23:37 |
dstufft | I know-ish erb | 23:38 |
fungi | it's erbalicious | 23:38 |
*** dteselkin has quit IRC | 23:38 | |
dstufft | perputual https://i.imgur.com/xVyoSl.jpg | 23:38 |
fungi | i feel like that every day | 23:38 |
greghaynes | but hey, you have a tie, thats kind of like knowing what youre doing | 23:38 |
nibalizer | oh the poodling? | 23:39 |
nibalizer | or depoodling | 23:39 |
fungi | yeah, the tie makes that photo | 23:39 |
*** sdake_ has joined #openstack-infra | 23:39 | |
fungi | yeah, i mixed up my vulnerability marketing buzzword names. i can't keep any of them straight | 23:39 |
dstufft | fungi: Does openstackid care about IE8 on Win XP (Python doesn't so we get moar better settings, but if openstackid.org does that's ok) | 23:40 |
*** katyafervent has joined #openstack-infra | 23:41 | |
fungi | dstufft: that's a great question. i don't personally care about those, but mrmartin and smarcet would be better to ask when they're around | 23:41 |
dstufft | specifically it's not all IE8, just IE8 on Win XP (although it is *all* IEs on WinXP), I'll include compat for it for now and if people don't care if it's lost it can be dropped easily | 23:42 |
*** katyafervent has quit IRC | 23:42 | |
*** rakhmerov has joined #openstack-infra | 23:43 | |
fungi | hrm, so the apache config is definitely set to _use_ a chain cert and the file it mentions exists and is readable. still digging | 23:43 |
*** sdake_ has quit IRC | 23:44 | |
*** katyafervent has joined #openstack-infra | 23:44 | |
*** dteselkin has joined #openstack-infra | 23:45 | |
*** banix has joined #openstack-infra | 23:45 | |
*** ociuhandu has quit IRC | 23:45 | |
*** ayoung has joined #openstack-infra | 23:46 | |
*** tsufiev_ has joined #openstack-infra | 23:46 | |
*** zz_dimtruck is now known as dimtruck | 23:48 | |
*** carl_baldwin has quit IRC | 23:48 | |
fungi | weird. it's set to use the same chain cert as all our other sites which are signed by the same ca | 23:49 |
clarkb | fungi: is the file corrupt somehow? | 23:49 |
clarkb | and openssl/apache lolnope without screaming too hard? | 23:50 |
fungi | md5sum is identical for /etc/ssl/certs/intermediate.pem on openstackid.org vs openstackid-dev.openstack.org | 23:50 |
*** Sukhdev has quit IRC | 23:51 | |
*** gordc has quit IRC | 23:51 | |
openstackgerrit | Donald Stufft proposed openstack-infra/puppet-openstackid: Use a set of secure ciphers, prefering PFS, AESGCM, AES, and then 3DES https://review.openstack.org/163228 | 23:52 |
dstufft | fungi: there you go ^ | 23:52 |
clarkb | 3DES is secure? | 23:52 |
dstufft | yup | 23:52 |
dstufft | it's secure but slow | 23:52 |
clarkb | I thought it wasn't | 23:52 |
dstufft | DES isn't | 23:52 |
dstufft | 3DES is | 23:52 |
*** otter768 has joined #openstack-infra | 23:52 | |
dstufft | it's also the only cipher that IE8 on Windows XP supports that isn't RC4 or an export cipher | 23:53 |
dstufft | if you don't care about IE8 on WinXP you can drop 3DES, it only exists to maintain compat with that | 23:53 |
anteaya | jeblair tristanC sorry I had to get out for a walk while there was still some sunshine and I'm glad I did | 23:54 |
dstufft | https://www.ssllabs.com/ssltest/viewClient.html?name=IE&version=8&platform=XP | 23:54 |
anteaya | I'm glad to see tristanC embracing his role as election official and hearing opinions and making decisions that he can live with | 23:54 |
anteaya | tristanC: well done | 23:54 |
*** chlong has joined #openstack-infra | 23:54 | |
clarkb | I am not sure we care about XP isn't XP dead? | 23:54 |
anteaya | jeblair: I do agree with you that the document is unclear | 23:54 |
clarkb | ya for almost a year now | 23:55 |
anteaya | I further agree that all documentation pertaining to tc governed election could stand some de-fogging | 23:55 |
*** yamamoto has joined #openstack-infra | 23:55 | |
*** gyee has quit IRC | 23:55 | |
anteaya | my preference is to propose changes to governance in milestone 1 of a realease when folks can think the most objectivly about the impact thereof | 23:55 |
dstufft | clarkb: want me to drop IE8/WinXP compat? I only included it because the old ciphers had support for it and I didn't want to drop support without someone telling me it's ok | 23:56 |
fungi | so according to openssl x509 -text -in intermediate.pem this is O=COMODO CA Limited, CN | 23:56 |
fungi | =PositiveSSL CA 2 | 23:56 |
jroll | 3DES sounds like a video game | 23:56 |
fungi | darn those stray line breaks | 23:56 |
anteaya | whereas milestone 3 tends to be a bad time for policy changes to election stuff if we can avoid it | 23:56 |
clarkb | dstufft: I think I would be fine with not explicitly supporting XP as that platform should be considered insecure at this point due to lack of support | 23:56 |
fungi | signed by O=AddTrust AB, OU=AddTrust External TTP Network, CN=AddTrust External CA Root | 23:56 |
greghaynes | Reguardless, RC4 is a whole other level of insecure so I think its safe to say the patch dstufft put up is a strict improvement | 23:56 |
*** dimtruck is now known as zz_dimtruck | 23:57 | |
jroll | I wonder if there's any large corps participating in openstack that still run xp | 23:57 |
anteaya | I also had the time to think about a possible direction for scaling tooling at least for the self-nomination portion | 23:57 |
*** otter768 has quit IRC | 23:57 | |
*** hdd_ has joined #openstack-infra | 23:57 | |
anteaya | I'm going to fix some dinner and draw up an etherpad | 23:57 |
dstufft | Note: You can still use Windows XP, you just have to use chrome or firefox | 23:57 |
dstufft | if 3DES is dropped | 23:57 |
jroll | dstufft: I'd be shocked if a company required XP but let users install software | 23:58 |
dstufft | IE6 on Windows XP doens't work because no SSLv3, IE8 requires 3DES or RC4, RC4 is broken, 3DES is slow as molases (and it's the only cipher that doesn't have PFS in that cipher set, so dropping it means you promise all connections get PFS), IE9+ doesn't run on WinXP | 23:59 |
*** zul has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!