*** lvdongbing has quit IRC | 00:00 | |
jeblair | pabelanger, clarkb: oh i know what it is | 00:00 |
---|---|---|
jeblair | pabelanger, clarkb: those are used nodes that nodepool never saw as being used because nodepool does not listed to zmq from zlstatic01 | 00:01 |
*** shashank_hegde has joined #openstack-infra | 00:01 | |
pabelanger | sounds right | 00:01 |
jeblair | pabelanger: so i think you'll need to 'nodepool delete' all of the zlstatic01 nodes after it stops | 00:01 |
pabelanger | ya, I was planning on doing that just to be safe | 00:01 |
clarkb | gotcha, zlstatic is like jenkins.o.o was and has no dynamic hosts | 00:01 |
pabelanger | if zuul didn't do it | 00:02 |
jeblair | yep. normally, except for that bug. | 00:02 |
*** garyk has quit IRC | 00:02 | |
clarkb | completely unrelated but we should totally have a grep -i error consolelog && exit 1 build step | 00:03 |
jeblair | clarkb: we *could* mix static/dynamic, but it didn't seem worth changing how we do things. | 00:03 |
jeblair | (release.slave.o.o can still only be attached to one launcher, so it may as well be a special launcher) | 00:04 |
pabelanger | So, feature request, not sure how to design it. Some how list which zuul-launcher a job is running on from status.o.o/zuul page. Before we could hover over URLs to see the jenkins hostname in the console URL. Maybe a enable a tooltip? | 00:04 |
clarkb | sphinx builds in aprticular complain about a lot of stuff but seem to build anyways and tox returns 0 | 00:04 |
*** garyk has joined #openstack-infra | 00:04 | |
jeblair | pabelanger: i hesitate for that -- i'm not sure which zuul-launcher is particularly useful feature to most users | 00:04 |
jeblair | clarkb: ugh, i thought we managed to convince sphinx to exit on warnings | 00:05 |
jeblair | error on warnings | 00:05 |
clarkb | jeblair: doesn't look like it | 00:05 |
pabelanger | jeblair: right, was looking for a easy way to see how many jobs were running when we do maintenance | 00:05 |
clarkb | jeblair: but these are even errors not just warnings | 00:05 |
pabelanger | doesn't have to be status.o.o | 00:05 |
jeblair | pabelanger: i run 'ps' on the launcher in question | 00:05 |
jeblair | pabelanger: you can see the 'ansible-playbook' processes | 00:06 |
pabelanger | ya, I do that too | 00:06 |
jeblair | pabelanger: i think there are 3 processes for each job | 00:06 |
jeblair | clarkb: wow :( | 00:06 |
clarkb | I am going to look in some zuul side logs to see if they persist there | 00:07 |
*** garyk has quit IRC | 00:07 | |
*** heyongli has quit IRC | 00:07 | |
clarkb | /home/jenkins/openstack/ceilometer/doc/source/webapi/v2.rst:234: ERROR: Unexpected indentation. | 00:07 |
clarkb | is one example | 00:07 |
*** heyongli has joined #openstack-infra | 00:07 | |
clarkb | http://logs.openstack.org/75/338975/2/check/gate-ceilometer-docs/b00c023/console.html.gz#_2016-07-07_14_57_47_839664 ya | 00:08 |
jeblair | clarkb: grepping in the console log is a neat idea, but i worry that makes us un-future-zuulv3 proof | 00:08 |
jeblair | clarkb: it would probably be better to fix this in the tox docs script | 00:08 |
*** thiagop has quit IRC | 00:09 | |
clarkb | does zuulv3 not have console logs? | 00:09 |
clarkb | but yes fixing it more directly is always better | 00:09 |
clarkb | can give better error messages to devs that way | 00:09 |
*** esberglu has joined #openstack-infra | 00:09 | |
jeblair | clarkb: i have no idea if zuulv3 will have console logs :) | 00:09 |
jeblair | i hope it has something like them | 00:10 |
jeblair | but how we end up doing some of this with ansible is still a bit of a wildcard | 00:10 |
*** garyk has joined #openstack-infra | 00:10 | |
clarkb | I see | 00:10 |
openstackgerrit | Merged openstack-infra/git-review: Update gerrit version for testing https://review.openstack.org/282018 | 00:10 |
clarkb | (and to be extra doubly clear that ERROR above did not cause the job to error) | 00:10 |
pabelanger | sphinx and pbr, always in for a fun time | 00:11 |
pabelanger | warnonerrors was broken for a while | 00:11 |
pabelanger | and only recently got fixed again | 00:11 |
clarkb | pabelanger: pretty sure its not working now | 00:11 |
clarkb | based on ^ | 00:11 |
pabelanger | clarkb: yup, seems to look that way | 00:12 |
clarkb | or maybe ceilometer has it diasbled? | 00:12 |
pabelanger | that time I looked, there was very basic to no testing with sphinx in pbr | 00:12 |
pabelanger | that was 18months ago | 00:12 |
fungi | disabling that was one of the workarounds mentioned in the pbr breaking change thread on the ml, so entirely possible | 00:12 |
jeblair | the run-docs script is pretty simple -- the exit code (0 in this case) is the exit code of tox -e$venv -- python setup.py build_sphinx | 00:13 |
jeblair | ceilometer has warnerrors = true | 00:14 |
clarkb | jeblair: ya my mega loop is just tox -evenv python setup.py build_sphinx and check $? and record some data on failure | 00:15 |
clarkb | then I can run find logs/ -name '*fail*' and find all the failed tox targets | 00:16 |
clarkb | (I am reasonably happy with how well this simple test all the things setup is working) | 00:16 |
clarkb | cisco-ironic-contrib is failibng because nova changed some internal interfaces | 00:16 |
clarkb | I am discovering most of the work here is going to be sifting the legit xenial issues from the $random things is broken | 00:17 |
*** heyongli has quit IRC | 00:17 | |
*** heyongli has joined #openstack-infra | 00:17 | |
*** PollyZ has joined #openstack-infra | 00:18 | |
*** caowei has quit IRC | 00:18 | |
*** zz_dimtruck is now known as dimtruck | 00:19 | |
*** ddieterly has joined #openstack-infra | 00:19 | |
*** _oanson has quit IRC | 00:20 | |
*** karthikp_ has joined #openstack-infra | 00:21 | |
*** bapalm has quit IRC | 00:22 | |
*** PollyZ has quit IRC | 00:22 | |
*** Julien-zte has joined #openstack-infra | 00:23 | |
*** bapalm has joined #openstack-infra | 00:25 | |
*** xarses has quit IRC | 00:26 | |
*** heyongli has quit IRC | 00:27 | |
*** heyongli has joined #openstack-infra | 00:28 | |
*** rlandy has quit IRC | 00:28 | |
pabelanger | clarkb: nice work | 00:30 |
pabelanger | jeblair: do we feel ansible is stable enough for zuul to start introducing 331499? SSH pipelining | 00:32 |
*** sflanigan is now known as sflanigan_meetin | 00:33 | |
*** cody-somerville has quit IRC | 00:33 | |
*** adrian_otto has quit IRC | 00:33 | |
*** pvaneck has quit IRC | 00:33 | |
jeblair | pabelanger: what was the answer wrt async? if async doesn't use this, what's the point? | 00:34 |
*** karthikp_ has quit IRC | 00:35 | |
clarkb | coupler has no code in it | 00:35 |
pabelanger | jeblair: ya, it doesn't look like async support it. That was my impression looking at the code. However, I still think it would be useful relating to all other things, rsync included | 00:35 |
clarkb | but runs python-jobs | 00:35 |
*** karthikp_ has joined #openstack-infra | 00:35 | |
*** cody-somerville has joined #openstack-infra | 00:36 | |
clarkb | maybe we should send email to the coupler fokls and ask them if they are still intending to do stuff wit htaht repo and retire it fi not | 00:36 |
clarkb | https://review.openstack.org/#/c/302791/ is the change | 00:37 |
jeblair | pabelanger: i have to run, let's pick this up later | 00:37 |
pabelanger | jeblair: no rush | 00:37 |
clarkb | Rockyg: ^ submitter has a huawei addr maybe you know what is going on? | 00:37 |
*** heyongli has quit IRC | 00:38 | |
openstackgerrit | Merged openstack-infra/project-config: Update build-image.sh to reflect nodepool config https://review.openstack.org/340539 | 00:38 |
*** heyongli has joined #openstack-infra | 00:38 | |
*** gyee has quit IRC | 00:38 | |
Rockyg | gimme a sec, clarkb checking it out.... | 00:39 |
*** gfreeman has quit IRC | 00:39 | |
openstackgerrit | Merged openstack-infra/project-config: Fix name of layer build target https://review.openstack.org/340366 | 00:39 |
Rockyg | clarkb, I know th guy. english name of Howard. I'll sent hime an mail. He should b reading and on line, even if not on IRC | 00:41 |
clarkb | Rockyg: thank you | 00:43 |
*** sdake has quit IRC | 00:45 | |
*** shashank_hegde has quit IRC | 00:45 | |
dmsimard | Probably not the best place to ask, but -- in general, is there some sort of schedule for dot releases of supported releases ? (i.e, Mitaka) | 00:47 |
dmsimard | I looked at http://releases.openstack.org/mitaka/index.html and http://releases.openstack.org/ | 00:47 |
*** sputnik13 has joined #openstack-infra | 00:48 | |
*** heyongli has quit IRC | 00:48 | |
*** rdx_matt has joined #openstack-infra | 00:48 | |
*** shashank_hegde has joined #openstack-infra | 00:48 | |
*** heyongli has joined #openstack-infra | 00:48 | |
Rockyg | clarkb, sent an email, cc'ed you. He's generally very responsive. | 00:50 |
openstackgerrit | Merged openstack-infra/project-config: Add keystone v3 only nv job to designate https://review.openstack.org/340388 | 00:51 |
*** tphummel has quit IRC | 00:53 | |
*** rdx_matt has quit IRC | 00:53 | |
*** vasanthi has quit IRC | 00:53 | |
*** kzaitsev_mb has joined #openstack-infra | 00:54 | |
*** PollyZ has joined #openstack-infra | 00:54 | |
*** tphummel has joined #openstack-infra | 00:55 | |
clarkb | Rockyg: perfect, thanks again | 00:55 |
clarkb | dmsimard: dot releases? also you may have better luck in the releases or stable channel | 00:56 |
clarkb | dmsimard: oh dot releases I don't know why I kept reading that ad DoT eg department of transportation | 00:56 |
clarkb | dmsimard: they happen on a schedule set by the stable maint team iirc. | 00:57 |
dmsimard | clarkb: ok, I'll try and ask if their schedule is available somewhere. Thanks. | 00:57 |
clarkb | clearly I should call it a day | 00:57 |
*** heyongli has quit IRC | 00:58 | |
*** heyongli has joined #openstack-infra | 00:58 | |
*** Swami has quit IRC | 00:59 | |
*** PollyZ has quit IRC | 01:00 | |
*** sputnik13_ has joined #openstack-infra | 01:01 | |
*** garyk has quit IRC | 01:01 | |
*** cody-somerville has quit IRC | 01:02 | |
*** ddieterly has quit IRC | 01:02 | |
*** cody-somerville has joined #openstack-infra | 01:04 | |
*** cody-somerville has joined #openstack-infra | 01:04 | |
*** sputnik13 has quit IRC | 01:04 | |
pabelanger | clarkb: how longs are the checks for wait_for now? | 01:06 |
clarkb | evoque fails pep8 and py27 and not just on xenial | 01:06 |
pabelanger | this is for the restart_zuul_launcher.yaml playbook | 01:06 |
pabelanger | 2016-07-12 01:01:15,411 DEBUG zuul.LaunchServer: Stopped | 01:06 |
clarkb | pabelanger: the length is the asme as before we just use a single ssh connection intead of reconnecting on a poll interval | 01:06 |
pabelanger | but ansible wait_for is still waiting | 01:06 |
pabelanger | Hmm | 01:07 |
clarkb | pabelanger: it checks if the pid file goes away, double check that happened? | 01:07 |
pabelanger | clarkb: ya, pid is gone | 01:07 |
clarkb | pabelanger: are you running from puppetmaster? | 01:07 |
pabelanger | no, locally this time | 01:08 |
clarkb | (thinking maybe its an ansible version difference?) | 01:08 |
pabelanger | ya, let me double check that | 01:08 |
pabelanger | ansible-playbook 2.0.1.0 | 01:08 |
*** shashank_hegde has quit IRC | 01:08 | |
pabelanger | so, could be that | 01:08 |
pabelanger | okay, good enough for me | 01:08 |
*** heyongli has quit IRC | 01:08 | |
pabelanger | I'll stop my playbook and manually start it | 01:08 |
clarkb | http://docs.ansible.com/ansible/wait_for_module.html | 01:09 |
clarkb | the docs don't actually say what the checking interval is | 01:09 |
pabelanger | ya, I noticed that too | 01:09 |
clarkb | but it did work on all of the instances I ran it against when using ansible on the puppetmaster | 01:09 |
pabelanger | ack | 01:09 |
pabelanger | I'll look into it more tomorrow | 01:10 |
clarkb | ok let me know what you fnid. I am about to call it a day for real and go spend time with family | 01:10 |
pabelanger | yup, same | 01:10 |
clarkb | I just keep getting nerd sniped by all these failures | 01:10 |
clarkb | I keep worrying I have stumbled upon some major xenial issue but no its always projects that just don't work | 01:10 |
pabelanger | #status log zlstatic01.o.o back online | 01:11 |
openstackstatus | pabelanger: finished logging | 01:11 |
*** heyongli has joined #openstack-infra | 01:11 | |
*** sputnik13_ has quit IRC | 01:12 | |
*** zhurong has joined #openstack-infra | 01:12 | |
*** Apoorva_ has joined #openstack-infra | 01:13 | |
*** akshai has quit IRC | 01:13 | |
pabelanger | jhesketh: ianw: mind helping land https://review.openstack.org/#/c/340543/ | 01:13 |
*** kzaitsev_mb has quit IRC | 01:14 | |
ianw | pabelanger: ok, not keeping up ... mounted space not enough? | 01:14 |
pabelanger | ianw: incorrect path for job, permission denied | 01:15 |
clarkb | /opt should have more space fwiw. I too am curious what the motivation behind that is | 01:15 |
clarkb | +2 | 01:15 |
pabelanger | we also delete contents each build | 01:16 |
*** Apoorva has quit IRC | 01:16 | |
*** Apoorva_ has quit IRC | 01:17 | |
ianw | pabelanger: ahh, no sudo at this point? | 01:17 |
*** heyongli has quit IRC | 01:19 | |
*** heyongli has joined #openstack-infra | 01:19 | |
openstackgerrit | Joshua Hesketh proposed openstack-infra/release-tools: Automate retiring old branches https://review.openstack.org/333875 | 01:20 |
jhesketh | pabelanger: looking | 01:20 |
*** mtanin___ has joined #openstack-infra | 01:21 | |
*** thorst has joined #openstack-infra | 01:21 | |
pabelanger | ianw: Ya | 01:21 |
*** mtanino has quit IRC | 01:22 | |
*** dimtruck is now known as zz_dimtruck | 01:24 | |
*** Qiming has joined #openstack-infra | 01:25 | |
*** heyongli has quit IRC | 01:29 | |
*** heyongli has joined #openstack-infra | 01:29 | |
*** xarses has joined #openstack-infra | 01:29 | |
openstackgerrit | Ian Wienand proposed openstack-infra/system-config: Make sure non-root can write to volume storage https://review.openstack.org/340638 | 01:30 |
ianw | pabelanger / clarkb : ^ might help. same thing broke diskimage-builder when i switched it to use /opt as temp space | 01:30 |
openstackgerrit | Merged openstack-infra/project-config: Update PACKAGE_DIR for reprepro-import builder https://review.openstack.org/340543 | 01:31 |
*** Daisy has joined #openstack-infra | 01:32 | |
*** Daisy has quit IRC | 01:33 | |
*** Daisy has joined #openstack-infra | 01:33 | |
*** karthikp_ has quit IRC | 01:35 | |
*** weshay has quit IRC | 01:37 | |
*** sputnik13 has joined #openstack-infra | 01:38 | |
*** sflanigan_meetin is now known as sflanigan | 01:38 | |
*** heyongli has quit IRC | 01:39 | |
*** heyongli has joined #openstack-infra | 01:40 | |
*** jaosorior has joined #openstack-infra | 01:40 | |
*** sputnik13_ has joined #openstack-infra | 01:40 | |
*** kzaitsev_mb has joined #openstack-infra | 01:41 | |
*** ajmiller has quit IRC | 01:42 | |
*** ajmiller has joined #openstack-infra | 01:42 | |
*** sputnik13 has quit IRC | 01:43 | |
*** karthikp_ has joined #openstack-infra | 01:45 | |
*** sputnik13 has joined #openstack-infra | 01:46 | |
*** sputnik13_ has quit IRC | 01:46 | |
*** thorst has quit IRC | 01:47 | |
*** Daisy has quit IRC | 01:47 | |
*** thorst has joined #openstack-infra | 01:47 | |
openstackgerrit | Craige McWhirter proposed openstack-infra/infra-specs: Initial commit of the Pholio spec https://review.openstack.org/340641 | 01:48 |
*** rdx_matt has joined #openstack-infra | 01:49 | |
*** heyongli has quit IRC | 01:49 | |
*** zul has quit IRC | 01:50 | |
*** heyongli has joined #openstack-infra | 01:50 | |
*** jianghuaw has joined #openstack-infra | 01:51 | |
*** thorst has quit IRC | 01:51 | |
*** piet has quit IRC | 01:52 | |
*** rdx_matt has quit IRC | 01:53 | |
openstackgerrit | Merged openstack-infra/project-config: Add openstack/fuel-plugin-murano-tests project https://review.openstack.org/332151 | 01:55 |
*** thorst has joined #openstack-infra | 01:55 | |
*** zz_dimtruck is now known as dimtruck | 01:56 | |
*** thorst has quit IRC | 01:56 | |
*** thorst has joined #openstack-infra | 01:56 | |
jianghuaw | hi, how are the tempest tests divided to workers? I run devstack-gate in my test environment, but I noticed that the longest worker ran much longer than the shortest worker. e.g. 1hr40min v.s. 0hr52min. | 01:57 |
jianghuaw | Is there anyway to make it to be better balanced? | 01:59 |
jianghuaw | thanks. | 01:59 |
*** PollyZ has joined #openstack-infra | 01:59 | |
*** tphummel has quit IRC | 01:59 | |
*** heyongli has quit IRC | 02:00 | |
*** ajmiller has quit IRC | 02:00 | |
openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Start fleshing out ansible-launcher tests https://review.openstack.org/317482 | 02:00 |
openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Clean up results returned from ansible-launcher https://review.openstack.org/317481 | 02:00 |
openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add jjb and zmq to requirements https://review.openstack.org/316752 | 02:00 |
openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Prepare files for testing zuulv25 https://review.openstack.org/316753 | 02:00 |
*** ajmiller has joined #openstack-infra | 02:00 | |
*** heyongli has joined #openstack-infra | 02:00 | |
clarkb | jianghuaw: using preseeded subunit files | 02:00 |
*** heyongli has quit IRC | 02:00 | |
clarkb | jianghuaw: you could possibly pressed with that data too but I dont have all the details handy | 02:00 |
*** zul_ has joined #openstack-infra | 02:01 | |
*** thorst has quit IRC | 02:02 | |
*** zul has joined #openstack-infra | 02:02 | |
jianghuaw | clarkb: Could you point me where is "preseeded subunit files"? | 02:02 |
*** Daisy has joined #openstack-infra | 02:03 | |
*** PollyZ has quit IRC | 02:03 | |
openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Move the command processing into the launcher cmd https://review.openstack.org/331123 | 02:04 |
clarkb | jianghuaw: it happens in project-config/nodepool/elememts somehwere for us | 02:04 |
ianw | jianguaw: have a look around -> http://git.openstack.org/cgit/openstack-infra/devstack-gate/tree/devstack-vm-gate.sh#n701 | 02:06 |
ianw | jianghuaw: ^ | 02:06 |
*** csomerville has joined #openstack-infra | 02:06 | |
jianghuaw | Ah, yes. It is. | 02:07 |
jianghuaw | clarkb and ianw: Thanks very much. Very helpful. | 02:07 |
*** sdake has joined #openstack-infra | 02:09 | |
*** cody-somerville has quit IRC | 02:10 | |
*** zul_ has quit IRC | 02:10 | |
*** kzaitsev_mb has quit IRC | 02:13 | |
*** Sukhdev_ has quit IRC | 02:15 | |
*** Daisy_ has joined #openstack-infra | 02:16 | |
*** Daisy has quit IRC | 02:17 | |
*** karthikp_ has quit IRC | 02:18 | |
*** karthikp_ has joined #openstack-infra | 02:19 | |
openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Give jenkins permission to read packaging_keytab https://review.openstack.org/340646 | 02:21 |
*** EricGonczer_ has joined #openstack-infra | 02:21 | |
*** rockstar_ has joined #openstack-infra | 02:24 | |
*** rkukura has quit IRC | 02:26 | |
*** sputnik13 has quit IRC | 02:27 | |
*** rockstar has quit IRC | 02:28 | |
*** yanyanhu has joined #openstack-infra | 02:32 | |
*** yanyanhu_ has joined #openstack-infra | 02:32 | |
*** yanyanhu_ has quit IRC | 02:32 | |
*** gongysh has joined #openstack-infra | 02:34 | |
*** amotoki has joined #openstack-infra | 02:36 | |
*** rfolco has quit IRC | 02:39 | |
ianw | ^ pabelanger : doesn't that mean you could "cat" it, and ... | 02:40 |
*** yamahata has quit IRC | 02:40 | |
*** rossella_s has quit IRC | 02:40 | |
*** rossella_s has joined #openstack-infra | 02:40 | |
*** kzaitsev_mb has joined #openstack-infra | 02:40 | |
*** piet has joined #openstack-infra | 02:43 | |
*** mtanino has joined #openstack-infra | 02:45 | |
*** mtanin___ has quit IRC | 02:47 | |
*** EricGonczer_ has quit IRC | 02:47 | |
*** rkukura has joined #openstack-infra | 02:49 | |
*** rdx_matt has joined #openstack-infra | 02:50 | |
*** mtanino has quit IRC | 02:50 | |
*** krtaylor has joined #openstack-infra | 02:50 | |
openstackgerrit | Sachi King proposed openstack-infra/zuul: Support post jobs by supporting rev checkout https://review.openstack.org/293194 | 02:52 |
*** piet has quit IRC | 02:54 | |
*** rdx_matt has quit IRC | 02:54 | |
*** piet has joined #openstack-infra | 02:54 | |
*** EricGonczer_ has joined #openstack-infra | 02:55 | |
*** Daisy_ has quit IRC | 02:57 | |
*** Daisy has joined #openstack-infra | 02:57 | |
*** sarob has joined #openstack-infra | 02:57 | |
*** amotoki has quit IRC | 02:59 | |
openstackgerrit | Weiwei Gu proposed openstack-dev/pbr: Fix argument order for assertEqual to (expected, observed) https://review.openstack.org/340654 | 03:00 |
*** ddieterly has joined #openstack-infra | 03:01 | |
*** Daisy has quit IRC | 03:02 | |
*** caowei has joined #openstack-infra | 03:02 | |
*** kzaitsev_mb has quit IRC | 03:10 | |
*** zhurong has quit IRC | 03:11 | |
*** Daisy has joined #openstack-infra | 03:12 | |
*** zhurong has joined #openstack-infra | 03:12 | |
*** sflanigan is now known as sflanigan_meetin | 03:12 | |
*** sflanigan_meetin is now known as sflanigan_afk | 03:12 | |
pabelanger | ianw: it does, but that's how we do it for afsadmin.keytab too | 03:13 |
pabelanger | ianw: but it runs on our static workers, so we have somewhat control on what can touch it | 03:13 |
pabelanger | http://logs.openstack.org/20/340520/4/experimental/reprepro-import-debian-openstack-jessie/109caed/console.html | 03:14 |
pabelanger | with it working properly | 03:14 |
pabelanger | ianw: jhesketh if you can help land 340646 I would appreciate it | 03:14 |
*** ddieterly has quit IRC | 03:15 | |
openstackgerrit | Eric Brown proposed openstack-infra/project-config: Enable py35 voting on oslo.config https://review.openstack.org/340657 | 03:15 |
*** ajmiller has quit IRC | 03:16 | |
*** ajmiller has joined #openstack-infra | 03:16 | |
*** piet has quit IRC | 03:16 | |
openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: WIP: ZooKeeper API https://review.openstack.org/340658 | 03:17 |
*** treaki_ has joined #openstack-infra | 03:18 | |
*** amotoki has joined #openstack-infra | 03:21 | |
*** treaki__ has quit IRC | 03:22 | |
openstackgerrit | Eric Brown proposed openstack-infra/project-config: Enable py35 voting on oslo.utils https://review.openstack.org/340662 | 03:23 |
*** ramishra has quit IRC | 03:24 | |
*** EricGonczer_ has quit IRC | 03:25 | |
*** amotoki has quit IRC | 03:26 | |
*** ramishra has joined #openstack-infra | 03:26 | |
*** ajmiller has quit IRC | 03:26 | |
*** ajmiller has joined #openstack-infra | 03:27 | |
*** yamamoto has quit IRC | 03:27 | |
*** yamamoto has joined #openstack-infra | 03:27 | |
*** amotoki has joined #openstack-infra | 03:27 | |
*** shashank_hegde has joined #openstack-infra | 03:27 | |
*** yamahata has joined #openstack-infra | 03:28 | |
*** armax has quit IRC | 03:29 | |
*** markvoelker has joined #openstack-infra | 03:32 | |
*** Daisy has quit IRC | 03:32 | |
*** Daisy has joined #openstack-infra | 03:33 | |
*** garyk has joined #openstack-infra | 03:34 | |
*** markvoelker_ has joined #openstack-infra | 03:35 | |
*** Daisy has quit IRC | 03:37 | |
*** shashank_hegde has quit IRC | 03:37 | |
*** shashank_hegde has joined #openstack-infra | 03:37 | |
*** kushal has joined #openstack-infra | 03:37 | |
*** markvoelker has quit IRC | 03:38 | |
*** sflanigan_afk is now known as sflanigan | 03:40 | |
*** karthikp_ has quit IRC | 03:43 | |
*** gouthamr has joined #openstack-infra | 03:44 | |
*** PollyZ has joined #openstack-infra | 03:45 | |
*** markvoelker has joined #openstack-infra | 03:46 | |
*** gouthamr has quit IRC | 03:49 | |
*** amotoki has quit IRC | 03:49 | |
*** markvoelker_ has quit IRC | 03:49 | |
*** shashank_hegde has quit IRC | 03:50 | |
*** rdx_matt has joined #openstack-infra | 03:50 | |
*** mtanino has joined #openstack-infra | 03:51 | |
*** _sarob has joined #openstack-infra | 03:52 | |
*** sarob has quit IRC | 03:53 | |
*** rdx_matt has quit IRC | 03:55 | |
*** gouthamr_ has joined #openstack-infra | 03:56 | |
*** julim has quit IRC | 03:58 | |
*** amotoki has joined #openstack-infra | 03:59 | |
*** tqtran has quit IRC | 04:00 | |
*** shashank_hegde has joined #openstack-infra | 04:00 | |
*** links has joined #openstack-infra | 04:02 | |
*** ramishra has quit IRC | 04:04 | |
*** ajmiller has quit IRC | 04:05 | |
*** ajmiller has joined #openstack-infra | 04:05 | |
openstackgerrit | Ian Wienand proposed openstack-infra/project-config: Add Fedora 24 nodes https://review.openstack.org/340671 | 04:06 |
openstackgerrit | Ian Wienand proposed openstack-infra/project-config: Add experimental Fedora 24 job https://review.openstack.org/340672 | 04:06 |
*** ramishra has joined #openstack-infra | 04:07 | |
*** sree has joined #openstack-infra | 04:07 | |
*** flepied1 has joined #openstack-infra | 04:10 | |
*** flepied has quit IRC | 04:11 | |
*** Ravikiran_K has quit IRC | 04:13 | |
*** Sukhdev has joined #openstack-infra | 04:14 | |
*** shashank_hegde has quit IRC | 04:16 | |
*** ramishra has quit IRC | 04:19 | |
*** gongysh has quit IRC | 04:21 | |
*** ramishra has joined #openstack-infra | 04:21 | |
openstackgerrit | Khai Do proposed openstack-infra/puppet-gerrit: Refactor to accept java_home from parent modules https://review.openstack.org/331523 | 04:21 |
openstackgerrit | Jin Jing Lin proposed openstack-infra/project-config: Enable Neutron VPNaaS API CI tests https://review.openstack.org/337011 | 04:23 |
*** davidlenwell has quit IRC | 04:28 | |
*** dimtruck is now known as zz_dimtruck | 04:29 | |
*** ramishra has quit IRC | 04:31 | |
*** ramishra has joined #openstack-infra | 04:31 | |
*** coolsvap has joined #openstack-infra | 04:37 | |
*** davidlenwell has joined #openstack-infra | 04:38 | |
clarkb | jacket is another project with no code | 04:41 |
*** ramishra has quit IRC | 04:41 | |
*** gouthamr_ has quit IRC | 04:42 | |
*** ramishra has joined #openstack-infra | 04:43 | |
*** _sarob has quit IRC | 04:46 | |
*** sarob has joined #openstack-infra | 04:46 | |
*** pahuang_ has quit IRC | 04:51 | |
*** rdx_matt has joined #openstack-infra | 04:51 | |
*** pahuang has joined #openstack-infra | 04:53 | |
*** rdx_matt has quit IRC | 04:56 | |
*** psachin has joined #openstack-infra | 04:56 | |
*** caowei has quit IRC | 04:56 | |
*** xyang1 has quit IRC | 04:58 | |
*** tqtran has joined #openstack-infra | 04:59 | |
*** shashank_hegde has joined #openstack-infra | 04:59 | |
*** ramishra has quit IRC | 05:00 | |
*** d0ugal has quit IRC | 05:01 | |
*** fawadkhaliq has joined #openstack-infra | 05:01 | |
*** ramishra has joined #openstack-infra | 05:01 | |
*** flepied1 has quit IRC | 05:02 | |
*** tqtran has quit IRC | 05:03 | |
*** shashank_hegde has quit IRC | 05:04 | |
*** caowei has joined #openstack-infra | 05:05 | |
*** Daisy has joined #openstack-infra | 05:05 | |
*** flepied has joined #openstack-infra | 05:07 | |
*** kzaitsev_mb has joined #openstack-infra | 05:07 | |
*** Daisy has quit IRC | 05:07 | |
*** flepied has quit IRC | 05:08 | |
*** hurgleburgler has joined #openstack-infra | 05:09 | |
*** sree has quit IRC | 05:10 | |
*** Daisy has joined #openstack-infra | 05:10 | |
*** shashank_hegde has joined #openstack-infra | 05:12 | |
*** kzaitsev_mb has quit IRC | 05:12 | |
*** caowei has quit IRC | 05:13 | |
*** sree has joined #openstack-infra | 05:14 | |
*** markvoelker_ has joined #openstack-infra | 05:16 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Change mistral-extra jobs https://review.openstack.org/340346 | 05:16 |
zaro | i made my change (https://review.openstack.org/#/c/331523/) depend on your change (https://review.openstack.org/#/c/340529/) which you thought might fix the lint puppet-lint job on my change but it doesn't seem to fix. would this be considered a satisfactory test that your change doesn't fix what you thought it might? | 05:18 |
*** zeih has quit IRC | 05:18 | |
*** david-lyle has joined #openstack-infra | 05:18 | |
*** markvoelker has quit IRC | 05:19 | |
zaro | clarkb: would be nice to cleanup gerrit once in a while. | 05:19 |
*** ajmiller has quit IRC | 05:21 | |
*** ajmiller has joined #openstack-infra | 05:22 | |
*** sandanar_ has joined #openstack-infra | 05:24 | |
*** sandanar has joined #openstack-infra | 05:24 | |
*** sdake_ has joined #openstack-infra | 05:25 | |
*** Daisy has quit IRC | 05:29 | |
*** Daisy has joined #openstack-infra | 05:29 | |
*** Daisy_ has joined #openstack-infra | 05:31 | |
*** Daisy has quit IRC | 05:31 | |
*** yanyanhu has quit IRC | 05:32 | |
*** Daisy_ has quit IRC | 05:33 | |
*** Daisy has joined #openstack-infra | 05:34 | |
jesusaur | zaro: no, the tests always use master for openstack_infra_spec_helper, so depending on the change doesn't have any effect :( | 05:34 |
*** hurgleburgler has quit IRC | 05:34 | |
*** ramishra has quit IRC | 05:35 | |
jesusaur | I tried looking into how to use a zuul ref for the spec helper, but it looks like bundler doesn't support that kind of thing | 05:35 |
sdake_ | AJaeger ping | 05:35 |
sdake_ | AJaeger i think i messed up your recheck of the ceph gate | 05:35 |
*** isaacb has joined #openstack-infra | 05:35 | |
*** savihou has joined #openstack-infra | 05:36 | |
openstackgerrit | Khai Do proposed openstack-infra/puppet-gerrit: Refactor to accept java_home from parent modules https://review.openstack.org/331523 | 05:38 |
*** itisha has quit IRC | 05:40 | |
*** esberglu has quit IRC | 05:41 | |
*** savihou has quit IRC | 05:41 | |
*** savihou has joined #openstack-infra | 05:42 | |
*** arif-ali has quit IRC | 05:45 | |
*** markvoelker_ has quit IRC | 05:45 | |
*** PollyZ has quit IRC | 05:46 | |
*** coolsvap has quit IRC | 05:46 | |
clarkb | zaro: ya I am going to script up a check for empty repos tomorrow I think and ping thwir owners | 05:48 |
*** hashar has joined #openstack-infra | 05:52 | |
*** rdx_matt has joined #openstack-infra | 05:52 | |
*** Sukhdev has quit IRC | 05:53 | |
sdake_ | AJaeger i can tell you rax.ord does not have the sources.available.d dir | 05:53 |
*** binbincong has quit IRC | 05:53 | |
sdake_ | pabelanger looked in one on of the images earlier and it had it on a different cloud provider | 05:53 |
*** Ravikiran_K has joined #openstack-infra | 05:54 | |
*** javeriak has joined #openstack-infra | 05:55 | |
*** sarob has quit IRC | 05:55 | |
sdake_ | clarkb any idea on the sources.d thing - been nt workign for 3+ days | 05:55 |
sdake_ | not in a uper rush | 05:56 |
sdake_ | but just want to make sure its working before i add more mirrors | 05:56 |
sdake_ | nt/not | 05:56 |
*** rdx_matt has quit IRC | 05:57 | |
*** fawadkhaliq has quit IRC | 05:58 | |
*** rcernin has joined #openstack-infra | 05:58 | |
*** Rockyg has quit IRC | 05:59 | |
*** rkukura has quit IRC | 06:00 | |
*** gongysh has joined #openstack-infra | 06:03 | |
*** amotoki has quit IRC | 06:03 | |
*** d0ugal has joined #openstack-infra | 06:04 | |
*** binbincong has joined #openstack-infra | 06:08 | |
*** rkukura has joined #openstack-infra | 06:10 | |
*** pgadiya has joined #openstack-infra | 06:10 | |
*** lennyb has quit IRC | 06:12 | |
*** rajinir has quit IRC | 06:15 | |
*** isaacb has quit IRC | 06:17 | |
*** amotoki has joined #openstack-infra | 06:17 | |
*** shardy has joined #openstack-infra | 06:17 | |
openstackgerrit | Merged openstack-infra/project-config: Add an experimental grenade multinode job for Neutron Linuxbridge https://review.openstack.org/336793 | 06:17 |
*** ramishra has joined #openstack-infra | 06:19 | |
*** lennyb has joined #openstack-infra | 06:20 | |
*** sdake has quit IRC | 06:22 | |
*** Daisy_ has joined #openstack-infra | 06:23 | |
*** mtanino has quit IRC | 06:24 | |
*** Daisy has quit IRC | 06:26 | |
lennyb | gmann: Could you, pls, review #link https://review.openstack.org/#/c/335447 | 06:28 |
*** pahuang has quit IRC | 06:30 | |
*** ramishra has quit IRC | 06:30 | |
*** ramishra has joined #openstack-infra | 06:31 | |
*** sree has quit IRC | 06:34 | |
*** sree has joined #openstack-infra | 06:35 | |
openstackgerrit | Eric Brown proposed openstack-infra/project-config: Enable py35 voting on vmware-nsx https://review.openstack.org/340708 | 06:35 |
*** nadya has joined #openstack-infra | 06:40 | |
*** pahuang has joined #openstack-infra | 06:42 | |
*** d0ugal has quit IRC | 06:43 | |
*** mahatic has quit IRC | 06:45 | |
*** nadya has quit IRC | 06:45 | |
*** csomerville has quit IRC | 06:46 | |
*** zeih has joined #openstack-infra | 06:46 | |
*** slaweq has joined #openstack-infra | 06:48 | |
*** caowei has joined #openstack-infra | 06:48 | |
*** sdake_ has quit IRC | 06:49 | |
*** yolanda has quit IRC | 06:50 | |
*** mahatic has joined #openstack-infra | 06:50 | |
*** isaacb has joined #openstack-infra | 06:50 | |
*** yolanda has joined #openstack-infra | 06:50 | |
*** nmagnezi has joined #openstack-infra | 06:50 | |
*** andreas_s has joined #openstack-infra | 06:50 | |
*** jswarren_ has quit IRC | 06:52 | |
*** amotoki has quit IRC | 06:52 | |
*** rdx_matt has joined #openstack-infra | 06:53 | |
openstackgerrit | Steven Hardy proposed openstack-infra/tripleo-ci: Temporarily pin puppet-ceph to stable/hammer https://review.openstack.org/340717 | 06:53 |
*** yolanda has quit IRC | 06:55 | |
*** fawadkhaliq has joined #openstack-infra | 06:56 | |
*** yanyanhu has joined #openstack-infra | 06:57 | |
*** rdx_matt has quit IRC | 06:58 | |
*** fawadkhaliq has quit IRC | 06:59 | |
*** fawadkhaliq has joined #openstack-infra | 06:59 | |
jamespage | kzaitsev_ws, yes - coreycb or I are good for ubuntu package queries | 06:59 |
*** andymaier has joined #openstack-infra | 06:59 | |
*** tesseract- has joined #openstack-infra | 07:00 | |
*** pcaruana has joined #openstack-infra | 07:00 | |
*** shardy has quit IRC | 07:01 | |
*** tqtran has joined #openstack-infra | 07:01 | |
*** yolanda has joined #openstack-infra | 07:03 | |
yolanda | good morning | 07:03 |
*** liusheng has joined #openstack-infra | 07:04 | |
*** ociuhandu has quit IRC | 07:06 | |
*** tqtran has quit IRC | 07:06 | |
*** ljjjustin has joined #openstack-infra | 07:08 | |
*** jianghuaw has quit IRC | 07:10 | |
*** yanyanhu_ has joined #openstack-infra | 07:12 | |
*** ramishra has quit IRC | 07:13 | |
*** yanyanhu has quit IRC | 07:14 | |
*** amotoki has joined #openstack-infra | 07:14 | |
*** ramishra has joined #openstack-infra | 07:15 | |
*** karthikp_ has joined #openstack-infra | 07:16 | |
*** dtardivel has joined #openstack-infra | 07:16 | |
*** Nikolay_St has joined #openstack-infra | 07:18 | |
*** garyk has quit IRC | 07:19 | |
*** amotoki has quit IRC | 07:19 | |
*** karthikp_ has quit IRC | 07:20 | |
*** garyk has joined #openstack-infra | 07:21 | |
*** david-lyle has quit IRC | 07:22 | |
*** ifarkas has joined #openstack-infra | 07:23 | |
*** Daisy has joined #openstack-infra | 07:30 | |
*** Daisy_ has quit IRC | 07:33 | |
openstackgerrit | Merged openstack-infra/system-config: Give jenkins permission to read packaging_keytab https://review.openstack.org/340646 | 07:35 |
*** vincentll has joined #openstack-infra | 07:36 | |
*** Daisy has quit IRC | 07:36 | |
*** Daisy has joined #openstack-infra | 07:36 | |
openstackgerrit | Merged openstack-infra/project-config: Use template for python35 everywhere https://review.openstack.org/339769 | 07:40 |
*** jlanoux has joined #openstack-infra | 07:43 | |
*** zhurong has quit IRC | 07:46 | |
*** eezhova has joined #openstack-infra | 07:46 | |
openstackgerrit | Merged openstack-infra/project-config: Cleanup after ansible EOLing https://review.openstack.org/340384 | 07:46 |
*** zhurong has joined #openstack-infra | 07:47 | |
openstackgerrit | Merged openstack-infra/project-config: Remove neutron full job from neutron-vpnaas queues https://review.openstack.org/340571 | 07:48 |
*** flepied has joined #openstack-infra | 07:49 | |
openstackgerrit | Merged openstack-infra/project-config: make gate-vmware-nsx-python34 voting https://review.openstack.org/339724 | 07:49 |
openstackgerrit | Merged openstack-infra/project-config: Don't run gate-cinder-tox-db-functional on stable/liberty https://review.openstack.org/340604 | 07:51 |
*** dfflanders has quit IRC | 07:52 | |
*** daemontool has joined #openstack-infra | 07:53 | |
*** rdx_matt has joined #openstack-infra | 07:53 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Change mistral-extra jobs https://review.openstack.org/340346 | 07:54 |
*** mikelk has joined #openstack-infra | 07:54 | |
*** flepied has quit IRC | 07:54 | |
*** dizquierdo has joined #openstack-infra | 07:56 | |
*** pt_15 has quit IRC | 07:56 | |
*** zeih has quit IRC | 07:56 | |
*** mrmartin has joined #openstack-infra | 07:58 | |
*** ramishra has quit IRC | 07:58 | |
*** rdx_matt has quit IRC | 07:58 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Change mistral-extra jobs https://review.openstack.org/340346 | 07:58 |
*** coolsvap has joined #openstack-infra | 07:59 | |
*** ramishra has joined #openstack-infra | 07:59 | |
*** fawadkhaliq has quit IRC | 07:59 | |
*** zzzeek has quit IRC | 08:00 | |
*** zzzeek has joined #openstack-infra | 08:00 | |
*** arxcruz has joined #openstack-infra | 08:00 | |
*** zeih has joined #openstack-infra | 08:01 | |
*** fawadkhaliq has joined #openstack-infra | 08:01 | |
therve | We're having very poor connectivity to tarballs.openstack.org recently | 08:03 |
therve | I don't know where the problem comes from, but it got worse in the past week or so | 08:03 |
*** asettle has joined #openstack-infra | 08:04 | |
*** fawadkhaliq has quit IRC | 08:05 | |
*** sorantis has joined #openstack-infra | 08:06 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Change mistral-extra jobs https://review.openstack.org/340346 | 08:07 |
*** amotoki has joined #openstack-infra | 08:09 | |
*** kzaitsev_mb has joined #openstack-infra | 08:09 | |
*** john-davidge has joined #openstack-infra | 08:09 | |
*** john-dav_ has joined #openstack-infra | 08:11 | |
*** john-dav_ is now known as john-davidge_ | 08:11 | |
rakhmerov | hi, I'm trying to get this done: https://review.openstack.org/#/c/340346 | 08:11 |
rakhmerov | and seems like I need some help from your team | 08:11 |
rakhmerov | is anybody available? | 08:12 |
*** jlanoux has quit IRC | 08:13 | |
*** jlanoux has joined #openstack-infra | 08:13 | |
*** amotoki has quit IRC | 08:13 | |
*** jtomasek_ has joined #openstack-infra | 08:13 | |
*** john-davidge has quit IRC | 08:14 | |
yolanda | hi rakhmerov , project-config cores and infra cores go through review queue daily. Your change will be reviewed soon | 08:14 |
*** kzaitsev_mb has quit IRC | 08:15 | |
*** ramishra has quit IRC | 08:15 | |
*** lucas|afk is now known as lucasagomes | 08:15 | |
rakhmerov | yeah, I know you review quickly, but I wanted to save time and ask some basic questions to reduce overall merge time ) | 08:15 |
yolanda | what do you need? | 08:15 |
*** jpich has joined #openstack-infra | 08:15 | |
rakhmerov | yolanda: we have "mistral-extra" repo which is configured to run standard jobs (py27, py34, pep8), I want to disable all these jobs | 08:16 |
rakhmerov | and I don't understand what changes I need to do in jenkins/jobs/projects.yaml | 08:17 |
*** ramishra has joined #openstack-infra | 08:17 | |
rakhmerov | can I still leave something like python-jobs in "jobs" section or should I replace it with something else | 08:17 |
yolanda | you need to remove any specific job that affects only to mistra-extra if you have any, and in zuul/layout.yaml remove that jobs, and add just merge-check and noop-jobs | 08:17 |
*** gongysh has quit IRC | 08:17 | |
rakhmerov | valid changes in zuul/layout.yaml are well explained in the doc | 08:17 |
*** AJaeger has quit IRC | 08:18 | |
rakhmerov | zuul/layout.yaml is fine, I did it | 08:18 |
yolanda | rakhmerov, if you don't need any job in mistral-extra, just remove the entry from jenkins/jobs/projects.yaml | 08:18 |
rakhmerov | ooh, ok | 08:18 |
rakhmerov | that's what I need to do | 08:18 |
*** Na3iL has joined #openstack-infra | 08:19 | |
rakhmerov | yolanda: thanks a bunch! | 08:19 |
yolanda | rakhmerov, glad to help | 08:19 |
*** coolsvap_ has joined #openstack-infra | 08:19 | |
*** derekh has joined #openstack-infra | 08:20 | |
*** rdx_matt has joined #openstack-infra | 08:22 | |
*** jistr has joined #openstack-infra | 08:22 | |
*** coolsvap has quit IRC | 08:22 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Change mistral-extra jobs https://review.openstack.org/340346 | 08:22 |
*** zeih has quit IRC | 08:24 | |
*** zeih has joined #openstack-infra | 08:26 | |
*** shashank_hegde has quit IRC | 08:27 | |
openstackgerrit | Merged openstack-infra/tripleo-ci: Bump up the number of OVB env workers available https://review.openstack.org/340359 | 08:28 |
*** EricGonczer_ has joined #openstack-infra | 08:28 | |
*** snarwade has quit IRC | 08:34 | |
*** snarwade has joined #openstack-infra | 08:34 | |
*** watersoul has quit IRC | 08:35 | |
*** nadya has joined #openstack-infra | 08:36 | |
*** watersoul has joined #openstack-infra | 08:36 | |
*** ajmiller has quit IRC | 08:37 | |
*** sshnaidm|afk is now known as sshnaidm | 08:38 | |
*** ajmiller has joined #openstack-infra | 08:38 | |
*** pgadiya has quit IRC | 08:38 | |
*** pgadiya has joined #openstack-infra | 08:41 | |
*** Daisy_ has joined #openstack-infra | 08:43 | |
*** chem has joined #openstack-infra | 08:43 | |
*** Na3iL has quit IRC | 08:45 | |
*** coolsvap_ has quit IRC | 08:46 | |
*** alexey_weyl has joined #openstack-infra | 08:48 | |
*** Daisy has quit IRC | 08:48 | |
alexey_weyl | Hi, | 08:49 |
alexey_weyl | We are having a problem in Vitrage project in the tempests, and we think that there might have happened some changes in openstack that maybe causing that | 08:50 |
alexey_weyl | When we run the tempests we can see in the vitrage logs that: | 08:50 |
alexey_weyl | " | 08:50 |
alexey_weyl | "DiscoveryFailure: Could not determine a suitable URL for the plugin" happens | 08:50 |
*** e0ne has joined #openstack-infra | 08:50 | |
alexey_weyl | You can see that here: | 08:51 |
alexey_weyl | #link http://logs.openstack.org/75/340275/3/check/gate-vitrage-dsvm-api/17314e5/logs/screen-vitrage-graph.txt.gz | 08:51 |
alexey_weyl | at 2016-07-11 14:27:49.759 | 08:51 |
*** sree has quit IRC | 08:51 | |
alexey_weyl | Maybe you are familiar with that problem? | 08:52 |
*** sorantis has quit IRC | 08:52 | |
*** sflanigan has quit IRC | 08:52 | |
*** javeriak has quit IRC | 08:54 | |
*** electrofelix has joined #openstack-infra | 08:56 | |
*** sorantis has joined #openstack-infra | 08:56 | |
*** pgadiya has quit IRC | 08:57 | |
*** idan_hefetz has joined #openstack-infra | 08:58 | |
*** pgadiya has joined #openstack-infra | 08:59 | |
*** dizquierdo has quit IRC | 09:00 | |
*** yamahata has quit IRC | 09:02 | |
*** fawadkhaliq has joined #openstack-infra | 09:02 | |
*** tqtran has joined #openstack-infra | 09:03 | |
*** amotoki has joined #openstack-infra | 09:03 | |
*** arxcruz has quit IRC | 09:03 | |
mordred | alexey_weyl: looks like an issue with your clinet connections and your auth parameters ... jamielennox ^^ ? yeah? | 09:05 |
*** tosky has joined #openstack-infra | 09:06 | |
*** fawadkhaliq has quit IRC | 09:06 | |
*** xarses_ has joined #openstack-infra | 09:07 | |
*** claudiub has joined #openstack-infra | 09:07 | |
*** tqtran has quit IRC | 09:07 | |
*** xarses has quit IRC | 09:07 | |
*** xarses_ has quit IRC | 09:07 | |
*** gongysh has joined #openstack-infra | 09:07 | |
*** xarses_ has joined #openstack-infra | 09:08 | |
*** dtantsur|afk is now known as dtantsur | 09:08 | |
*** sputnik13 has joined #openstack-infra | 09:08 | |
alexey_weyl | I agree, but I checked and saw that in Vitrage nothing has changed in that area, and thus I don't see why this is happening | 09:08 |
alexey_weyl | because until now it worked | 09:08 |
*** raildo-afk has quit IRC | 09:09 | |
*** amotoki has quit IRC | 09:09 | |
mordred | alexey_weyl: I haven't been following changes in this area recently... hopefully jamielennox has some thoughts or knows who to go poke | 09:10 |
*** Daisy_ has quit IRC | 09:10 | |
*** Daisy has joined #openstack-infra | 09:11 | |
*** EricGonczer_ has quit IRC | 09:12 | |
*** caowei has quit IRC | 09:13 | |
*** ramishra has quit IRC | 09:14 | |
*** ramishra has joined #openstack-infra | 09:15 | |
*** Daisy has quit IRC | 09:15 | |
*** arxcruz has joined #openstack-infra | 09:17 | |
openstackgerrit | Derek Higgins proposed openstack-infra/tripleo-ci: Temporarily pin puppet-ceph to stable/hammer https://review.openstack.org/340717 | 09:19 |
openstackgerrit | Oleksii Zamiatin proposed openstack-infra/project-config: Turn zmq gates to voting https://review.openstack.org/340439 | 09:20 |
*** yamahata has joined #openstack-infra | 09:22 | |
*** caowei has joined #openstack-infra | 09:23 | |
*** fawadkhaliq has joined #openstack-infra | 09:24 | |
*** ramishra has quit IRC | 09:25 | |
*** binbincong has quit IRC | 09:26 | |
*** ramishra has joined #openstack-infra | 09:26 | |
*** _degorenko|afk is now known as degorenko | 09:27 | |
*** binbincong has joined #openstack-infra | 09:28 | |
*** ccamacho has joined #openstack-infra | 09:28 | |
openstackgerrit | ChangBo Guo(gcb) proposed openstack-dev/pbr: Add Python 3.5 classifier and venv https://review.openstack.org/340813 | 09:28 |
openstackgerrit | Vladimir Kozhukalov proposed openstack-infra/project-config: Disable fuel-mirror pep8 jobs for master. https://review.openstack.org/340814 | 09:28 |
openstackgerrit | Oleksii Zamiatin proposed openstack-infra/project-config: [zmq] Use proper driver name to configure devstack https://review.openstack.org/339050 | 09:29 |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 09:30 |
openstackgerrit | Vladimir Kozhukalov proposed openstack-infra/project-config: Disable all gate jobs for fuel-mirror master https://review.openstack.org/340814 | 09:30 |
*** caowei has quit IRC | 09:32 | |
*** kzaitsev_mb has joined #openstack-infra | 09:35 | |
openstackgerrit | Merged openstack-infra/project-config: Add a bandit non-voting for Kolla https://review.openstack.org/290070 | 09:35 |
*** zeih has quit IRC | 09:36 | |
*** shardy has joined #openstack-infra | 09:40 | |
*** ramishra has quit IRC | 09:43 | |
*** ramishra has joined #openstack-infra | 09:46 | |
*** Guest2250 has quit IRC | 09:47 | |
*** jaosorior has quit IRC | 09:49 | |
*** jaosorior has joined #openstack-infra | 09:49 | |
*** gongysh has quit IRC | 09:53 | |
openstackgerrit | Derek Higgins proposed openstack-infra/tripleo-ci: Only run kill if there is a httpserver to kill https://review.openstack.org/340829 | 09:56 |
*** amotoki has joined #openstack-infra | 09:57 | |
*** ajmiller has quit IRC | 09:59 | |
*** ajmiller has joined #openstack-infra | 09:59 | |
*** javeriak has joined #openstack-infra | 10:00 | |
*** dizquierdo has joined #openstack-infra | 10:00 | |
*** d0ugal has joined #openstack-infra | 10:02 | |
*** amotoki has quit IRC | 10:02 | |
*** yamahata has quit IRC | 10:03 | |
jamielennox | alexey_weyl: there's a few things that could cause it, i would take the auth params and try them in osc with the --debug flag and see what its hitting | 10:07 |
jamielennox | alexey_weyl: the first thing i see as a problem is admin_endpoint in http://logs.openstack.org/75/340275/3/check/gate-vitrage-dsvm-api/17314e5/logs/etc/keystone/keystone.conf.txt.gz | 10:07 |
jamielennox | but i don't know if/why vitrage would be changing that from the standard devstack config | 10:08 |
*** eezhova has quit IRC | 10:09 | |
jamielennox | but i can't think of recent changes that would have messed with this from an auth end | 10:09 |
*** kzaitsev_mb has quit IRC | 10:10 | |
*** eezhova has joined #openstack-infra | 10:12 | |
*** zhurong has quit IRC | 10:13 | |
*** Julien-zte has quit IRC | 10:13 | |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 10:15 |
openstackgerrit | Marton Kiss proposed openstack-infra/groups: Add https support for mapbox urls https://review.openstack.org/340836 | 10:16 |
*** eezhova has quit IRC | 10:21 | |
*** yanyanhu_ has quit IRC | 10:24 | |
*** Julien-zte has joined #openstack-infra | 10:24 | |
*** sorantis has quit IRC | 10:24 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Add new "mistral-lib" project https://review.openstack.org/340844 | 10:25 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Remove tripleo-jobs.py https://review.openstack.org/337657 | 10:26 |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Change mistral-extra jobs https://review.openstack.org/340346 | 10:29 |
*** jlanoux_ has joined #openstack-infra | 10:29 | |
*** zeih has joined #openstack-infra | 10:30 | |
*** jlanoux has quit IRC | 10:30 | |
*** berendt has quit IRC | 10:33 | |
*** rossella_s has quit IRC | 10:40 | |
*** rossella_s has joined #openstack-infra | 10:40 | |
openstackgerrit | Renat Akhmerov proposed openstack-infra/project-config: Add new "mistral-lib" project https://review.openstack.org/340844 | 10:42 |
*** jlanoux_ has quit IRC | 10:44 | |
*** jlanoux has joined #openstack-infra | 10:44 | |
dulek | Hi everyone! We can notice some strange python jobs failures in Cinder gate: http://logs.openstack.org/80/337580/9/check/gate-cinder-python34-db/086b9af/console.html | 10:45 |
*** rodrigods has quit IRC | 10:48 | |
*** rodrigods has joined #openstack-infra | 10:48 | |
*** arif-ali has joined #openstack-infra | 10:50 | |
*** eezhova has joined #openstack-infra | 10:56 | |
*** apuimedo is now known as apuimedo|lunch | 10:57 | |
rcarrillocruz | mordred: https://review.openstack.org/#/c/315713/ fixes the func tests | 10:58 |
rcarrillocruz | i need to sort out why now the unit tests are failing (some mock related errors) and we should be good | 10:58 |
rcarrillocruz | Shrews: ^ | 10:58 |
rcarrillocruz | regardless, i really think we should get to the bottom of the admin vs non-admin performance issue | 11:02 |
rcarrillocruz | asselin__: ^ | 11:04 |
asselin__ | rcarrillocruz, will try it out | 11:05 |
*** julim has joined #openstack-infra | 11:05 | |
*** gyx has joined #openstack-infra | 11:06 | |
*** kzaitsev_mb has joined #openstack-infra | 11:06 | |
asselin__ | rcarrillocruz, were you able to reproduce the issue with clouds.yaml verify=false? | 11:07 |
*** sdake has joined #openstack-infra | 11:07 | |
rcarrillocruz | i haven't reached to that yet, being drowned by shade gate issues | 11:08 |
asselin__ | ok | 11:08 |
gyx | Is this temptest? | 11:08 |
*** sdake_ has joined #openstack-infra | 11:08 | |
asselin__ | rcarrillocruz, so how is the "ansible master" deployed? | 11:09 |
rcarrillocruz | it can be any machine with the ansible-role-puppet installed and some hiera | 11:09 |
rcarrillocruz | thing is, this was reused from puppetmaster | 11:10 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Remove Fedora support from CI scripts https://review.openstack.org/338210 | 11:10 |
rcarrillocruz | so it expects the same hiera datastructure | 11:10 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Remove pointless case clause https://review.openstack.org/338725 | 11:10 |
kzaitsev_mb | jamespage: zul: coreyob: is anyone of you guys the correct person to ask about ubuntu packages? | 11:11 |
*** zeih has quit IRC | 11:11 | |
jamespage | kzaitsev_mb, yes - ask away | 11:11 |
rcarrillocruz | one of the things i hope to do in infra-ansible | 11:11 |
rcarrillocruz | is to put puppetdb ( i have some wip around for puppet-openstackci ) | 11:11 |
rcarrillocruz | problem is, puppetdb requires ssl | 11:11 |
rcarrillocruz | and you have two options: | 11:11 |
rcarrillocruz | 1. install a puppetmaster to reuse its CA | 11:11 |
rcarrillocruz | 2. put puppetdb behind a ssl reverse proxy | 11:12 |
rcarrillocruz | in infra they just used 1, since they just reused puppetmaster | 11:12 |
rcarrillocruz | but for puppetopsntak ci it would be overkill | 11:12 |
*** berendt has joined #openstack-infra | 11:12 | |
rcarrillocruz | need to find some time to implement 2 | 11:12 |
asselin__ | agree | 11:12 |
*** sdake has quit IRC | 11:13 | |
*** Nikolay_St has quit IRC | 11:13 | |
*** rdxnoam has quit IRC | 11:14 | |
gyx | Is this the channel for temptest? | 11:15 |
*** mrmartin has quit IRC | 11:15 | |
*** Nikolay_St has joined #openstack-infra | 11:16 | |
rcarrillocruz | gyx: try better #openstack-qa | 11:16 |
*** winggundamth has quit IRC | 11:16 | |
openstackgerrit | Merged openstack-infra/groups: Add https support for mapbox urls https://review.openstack.org/340836 | 11:16 |
*** psilvad has quit IRC | 11:17 | |
*** gyx has left #openstack-infra | 11:17 | |
*** idan_hefetz has left #openstack-infra | 11:17 | |
openstackgerrit | Sam Betts proposed openstack-infra/project-config: Refactor Ironic job macro https://review.openstack.org/340567 | 11:18 |
*** tpsilva has joined #openstack-infra | 11:20 | |
*** zubchick has joined #openstack-infra | 11:22 | |
*** julim has quit IRC | 11:23 | |
*** amotoki has joined #openstack-infra | 11:25 | |
*** ddieterly has joined #openstack-infra | 11:26 | |
openstackgerrit | Graham Hayes proposed openstack-infra/project-config: Fix designate-identity-v3-only tempest job https://review.openstack.org/340878 | 11:29 |
*** ddieterly has quit IRC | 11:31 | |
*** watersoul has quit IRC | 11:31 | |
*** watersoul_ has joined #openstack-infra | 11:32 | |
*** dizquierdo has quit IRC | 11:33 | |
openstackgerrit | Xiang Chen proposed openstack/diskimage-builder: Fix a bug when using yum-minimal https://review.openstack.org/340884 | 11:35 |
zubchick | Hi folks. Is some environment variable exists that I can use in my tox.ini to determine branch name when commit/gate tests run on Jenkins? | 11:38 |
*** ldnunes has joined #openstack-infra | 11:39 | |
*** chem is now known as chem|away | 11:40 | |
*** thorst has joined #openstack-infra | 11:41 | |
*** weshay has joined #openstack-infra | 11:42 | |
*** sorantis has joined #openstack-infra | 11:43 | |
rcarrillocruz | ZUUL_BRANCH ? | 11:43 |
*** asettle has quit IRC | 11:44 | |
*** sambetts|afk is now known as sambetts | 11:44 | |
rcarrillocruz | but that's available on the test itself | 11:45 |
rcarrillocruz | not on tox.ini | 11:45 |
rcarrillocruz | not sure what you are after | 11:45 |
*** fawadkhaliq has quit IRC | 11:50 | |
*** rfolco has joined #openstack-infra | 11:51 | |
*** chem|away has quit IRC | 11:55 | |
*** jokke_ has quit IRC | 11:59 | |
*** asettle has joined #openstack-infra | 11:59 | |
*** mrmartin has joined #openstack-infra | 11:59 | |
*** jokke_ has joined #openstack-infra | 11:59 | |
*** markusry has joined #openstack-infra | 11:59 | |
*** Julien-zte has quit IRC | 12:03 | |
*** asettle has quit IRC | 12:04 | |
*** lucasagomes is now known as lucas-hungry | 12:04 | |
*** pradk has joined #openstack-infra | 12:04 | |
Zara | hm, trusty doesn't like https://review.openstack.org/#/c/330925/11 or https://review.openstack.org/#/c/340605/1 . seems to be the same sort of thing for both (only skimmed logs so far) | 12:05 |
*** tqtran has joined #openstack-infra | 12:05 | |
javeriak | hello, can someone please point me to the steps for tagging an openstack repo? | 12:05 |
*** ddieterly has joined #openstack-infra | 12:06 | |
*** Julien-zte has joined #openstack-infra | 12:06 | |
dtardivel | hi folks, could you help me about this changset https://review.openstack.org/#/c/327095/ which should be merged by Gerrit, and unfortunately I feel it's stuck somewhere in the gate :/ | 12:06 |
*** abregman has joined #openstack-infra | 12:07 | |
*** zeih has joined #openstack-infra | 12:08 | |
*** tqtran has quit IRC | 12:10 | |
*** abregman has quit IRC | 12:10 | |
*** abregman has joined #openstack-infra | 12:11 | |
*** amotoki_ has joined #openstack-infra | 12:13 | |
*** amotoki has quit IRC | 12:13 | |
*** TaylorHuang has joined #openstack-infra | 12:14 | |
*** ccamacho is now known as ccamacho|lunch | 12:15 | |
*** ajmiller has quit IRC | 12:16 | |
*** sdake_ has quit IRC | 12:16 | |
*** ajmiller has joined #openstack-infra | 12:16 | |
*** fawadkhaliq has joined #openstack-infra | 12:17 | |
*** TaylorH has quit IRC | 12:17 | |
*** dizquierdo has joined #openstack-infra | 12:17 | |
*** cbader has joined #openstack-infra | 12:19 | |
*** asettle has joined #openstack-infra | 12:20 | |
*** fawadkhaliq has quit IRC | 12:20 | |
*** akshai has joined #openstack-infra | 12:20 | |
*** piet has joined #openstack-infra | 12:22 | |
*** psachin has quit IRC | 12:24 | |
*** gouthamr has joined #openstack-infra | 12:24 | |
*** asettle has quit IRC | 12:25 | |
*** zeih has quit IRC | 12:25 | |
*** zhurong has joined #openstack-infra | 12:26 | |
*** trown|outtypewww is now known as trown | 12:26 | |
*** dprince has joined #openstack-infra | 12:27 | |
*** zeih has joined #openstack-infra | 12:27 | |
Zara | attempted to help debug gerrit-storyboard plugin config. I think https://review.openstack.org/#/c/330925/ is failing to find a certificate file or something, which seems familiar. I have no idea what's up with https://review.openstack.org/#/c/340605/ ; the error message cites a file which isn't changed as part of the patch. any ideas appreciated. :) | 12:28 |
*** Nikolay_St has quit IRC | 12:28 | |
*** mdrabe has joined #openstack-infra | 12:31 | |
*** ddieterly is now known as ddieterly[away] | 12:31 | |
*** ddieterly[away] is now known as ddieterly | 12:31 | |
*** ddieterly has quit IRC | 12:32 | |
*** kgiusti has joined #openstack-infra | 12:32 | |
*** sree has joined #openstack-infra | 12:32 | |
*** apuimedo|lunch is now known as apuimedo | 12:32 | |
*** akshai has quit IRC | 12:33 | |
* Zara finds https://review.openstack.org/#/c/331523/ | 12:33 | |
*** Julien-zte has quit IRC | 12:33 | |
*** markvoelker has joined #openstack-infra | 12:34 | |
Zara | I'm gonna hazard a guess that that's what the second patch was talking about... | 12:34 |
*** psachin has joined #openstack-infra | 12:34 | |
*** ddieterly has joined #openstack-infra | 12:35 | |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 12:36 |
*** psilvad has joined #openstack-infra | 12:37 | |
*** ihrachys has joined #openstack-infra | 12:37 | |
*** baoli has joined #openstack-infra | 12:39 | |
*** piet has quit IRC | 12:40 | |
*** piet has joined #openstack-infra | 12:40 | |
*** weshay is now known as weshay_mtg | 12:40 | |
*** amotoki_ has quit IRC | 12:41 | |
*** ddieterly has quit IRC | 12:42 | |
*** dizquierdo has quit IRC | 12:45 | |
*** Goneri has joined #openstack-infra | 12:46 | |
*** saggi1 has joined #openstack-infra | 12:46 | |
*** ccamacho|lunch is now known as ccamacho | 12:46 | |
*** john-davidge_ is now known as john-davidge | 12:47 | |
*** thiagop has joined #openstack-infra | 12:47 | |
*** rlandy has joined #openstack-infra | 12:47 | |
*** baoli_ has joined #openstack-infra | 12:47 | |
*** jed56 has joined #openstack-infra | 12:47 | |
*** burgerk has joined #openstack-infra | 12:47 | |
*** baoli has quit IRC | 12:49 | |
*** amotoki has joined #openstack-infra | 12:50 | |
*** saggi1 has quit IRC | 12:51 | |
*** mdrabe has quit IRC | 12:51 | |
*** eezhova has quit IRC | 12:51 | |
*** yaume has joined #openstack-infra | 12:52 | |
*** yanyanhu has joined #openstack-infra | 12:52 | |
*** ajmiller has quit IRC | 12:53 | |
*** ajmiller has joined #openstack-infra | 12:53 | |
*** yanyanhu has quit IRC | 12:53 | |
*** hashar_ has joined #openstack-infra | 12:53 | |
*** hashar__ has joined #openstack-infra | 12:53 | |
*** hashar_ has quit IRC | 12:54 | |
*** ramishra has quit IRC | 12:55 | |
*** lucas-hungry is now known as lucasagomes | 12:56 | |
*** yanyanhu has joined #openstack-infra | 12:56 | |
*** yanyanhu_ has joined #openstack-infra | 12:56 | |
*** eezhova has joined #openstack-infra | 12:56 | |
*** ramishra has joined #openstack-infra | 12:57 | |
*** M-docaedo_vector has quit IRC | 12:57 | |
*** edmondsw has joined #openstack-infra | 12:58 | |
*** javeriak has quit IRC | 12:58 | |
*** javeriak has joined #openstack-infra | 13:00 | |
*** xyang1 has joined #openstack-infra | 13:00 | |
*** makowals has joined #openstack-infra | 13:00 | |
odyssey4me | mgagne pabelanger ping? | 13:01 |
*** mdrabe has joined #openstack-infra | 13:01 | |
*** yanyanhu_ has left #openstack-infra | 13:01 | |
odyssey4me | javeriak http://docs.openstack.org/infra/manual/drivers.html#tagging-a-release | 13:02 |
*** amotoki has quit IRC | 13:03 | |
*** Na3iL has joined #openstack-infra | 13:03 | |
*** rdx_matt has quit IRC | 13:04 | |
*** woodster_ has joined #openstack-infra | 13:04 | |
*** rdx_matt has joined #openstack-infra | 13:04 | |
*** akshai has joined #openstack-infra | 13:04 | |
pabelanger | morning | 13:05 |
*** snarwade has quit IRC | 13:05 | |
pabelanger | odyssey4me: ^ | 13:05 |
lucasagomes | mordred, when trying to vote on the P release name I'm getting a "Your voter key is invalid." from the link in the email | 13:05 |
odyssey4me | morning pabelanger :) | 13:06 |
electrofelix | waynr: going to push up the rebased set of JJB 2.0 api patches later this afternoon as I'm away over the next few days so figure it would be good to allow others to re-review | 13:06 |
*** Hal has joined #openstack-infra | 13:07 | |
electrofelix | waynr: did you get a chance to look over the modifications I made to your patch queue and are reasonably happy with them? | 13:07 |
*** Hal is now known as Guest94106 | 13:07 | |
odyssey4me | pabelanger with reference to https://review.openstack.org/339770 I'd like to introduce an optional extension to the ansible role functional tests to allow the optional implementation of some sort of change in the test... and I realise now that it'll need a different name for the job | 13:07 |
*** Goneri has quit IRC | 13:07 | |
*** snarwade has joined #openstack-infra | 13:07 | |
*** alexey_weyl has quit IRC | 13:08 | |
*** rdx_matt has quit IRC | 13:08 | |
odyssey4me | pabelanger I was wondering whether we could change the job template to be '{pipeline}-{name}-ansible-{tox-environment}-{ostype}{job-suffix}' and have all the jobs set to use 'functional' as the tox-environment, unless the repo specifies an alternative | 13:09 |
odyssey4me | pabelanger as I recall we cut the name from functional to func due to job name length limitations - not sure if those are still in effect? | 13:09 |
*** Na3iL has quit IRC | 13:10 | |
*** Trozz has quit IRC | 13:11 | |
pabelanger | odyssey4me: I thought we had a {scenario} flag passed to the job name, so we could control the tox environment | 13:11 |
pabelanger | guess I was wrong | 13:11 |
odyssey4me | pabelanger yeah, that's kinda what I'd like to implement | 13:12 |
odyssey4me | with the default 'scenario' being 'functional' | 13:12 |
*** dizquierdo has joined #openstack-infra | 13:12 | |
pabelanger | odyssey4me: right, I am fine with that. Lets use scenario, like the puppet jobs and you are right, I think we are going to have some name length limitations | 13:13 |
pabelanger | odyssey4me: By brain is lacking coffee atm to help could up with a fix for that | 13:13 |
*** Na3iL has joined #openstack-infra | 13:13 | |
*** asettle has joined #openstack-infra | 13:13 | |
pabelanger | let me loop back to it | 13:13 |
odyssey4me | pabelanger alright, let me have a go at it while you come up with coffee | 13:14 |
*** ddieterly has joined #openstack-infra | 13:15 | |
*** Goneri has joined #openstack-infra | 13:16 | |
openstackgerrit | Ihar Hrachyshka proposed openstack-infra/project-config: Make gate-grenade-dsvm-neutron-dvr-multinode voting https://review.openstack.org/336116 | 13:16 |
javeriak | thanks odyssey4me :) | 13:16 |
openstackgerrit | Ihar Hrachyshka proposed openstack-infra/project-config: Make gate-grenade-dsvm-neutron-dvr-multinode voting https://review.openstack.org/336116 | 13:16 |
*** mriedem_away is now known as mriedem | 13:17 | |
openstackgerrit | Paul Belanger proposed openstack-infra/project-config: Fix URL variable with reprepro-import builder https://review.openstack.org/340937 | 13:18 |
*** esberglu has joined #openstack-infra | 13:19 | |
*** pgadiya has quit IRC | 13:20 | |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 13:20 |
*** pgadiya has joined #openstack-infra | 13:20 | |
*** pgadiya has quit IRC | 13:20 | |
*** M-docaedo_vector has joined #openstack-infra | 13:20 | |
*** ddieterly is now known as ddieterly[away] | 13:20 | |
*** julim has joined #openstack-infra | 13:21 | |
*** zeih has quit IRC | 13:21 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 13:22 |
*** cardeois has joined #openstack-infra | 13:22 | |
*** julim has quit IRC | 13:23 | |
*** amotoki has joined #openstack-infra | 13:23 | |
odyssey4me | pabelanger ^ not sure if that will work - I'm trying to allow the job name to remain short, but provide the option of the scenario setting | 13:23 |
*** yanyanhu has quit IRC | 13:24 | |
*** yanyanhu has joined #openstack-infra | 13:25 | |
*** zhurong has quit IRC | 13:25 | |
*** ddieterly[away] is now known as ddieterly | 13:26 | |
pabelanger | odyssey4me: https://review.openstack.org/#/c/287850/ is the issue with char limits | 13:27 |
*** tonytan4ever has joined #openstack-infra | 13:27 | |
*** jcoufal has joined #openstack-infra | 13:28 | |
openstackgerrit | Steven Hardy proposed openstack/diskimage-builder: DNM testing tripleo-ci fix https://review.openstack.org/340941 | 13:28 |
*** amotoki has quit IRC | 13:28 | |
pabelanger | odyssey4me: we should really dig in to tox and find out why 128 is the char limit | 13:29 |
john-davidge | anteaya: Hello, HP Networking CI has started posting a comment at the beginning of its test run. See https://review.openstack.org/#/c/337851 for an example | 13:29 |
*** jpich_ has joined #openstack-infra | 13:29 | |
odyssey4me | pabelanger as I recall it wasn't tox, it was jenkins | 13:29 |
odyssey4me | so maybe the issue is no more | 13:29 |
*** jtomasek__ has joined #openstack-infra | 13:29 | |
odyssey4me | maybe fungi can remember? | 13:29 |
pabelanger | odyssey4me: the issue is the python interpreter rewritten for python files in tox, explodes on long paths | 13:30 |
*** d0ugal_ has joined #openstack-infra | 13:30 | |
pabelanger | which result in python apps not being able to rung | 13:30 |
odyssey4me | oh, that's a whole different ball game | 13:31 |
*** pt_15 has joined #openstack-infra | 13:31 | |
Shrews | rcarrillocruz: really? interesting. notmorgan could not reproduce the issue on a new devstack setup yesterday. i'm thoroughly confused | 13:31 |
*** itisha has joined #openstack-infra | 13:31 | |
rcarrillocruz | well | 13:31 |
rcarrillocruz | the fix i referred | 13:31 |
odyssey4me | pabelanger hmm, does it have something to do with the home folder's name - or some other file name on the FS? | 13:31 |
rcarrillocruz | is not related to the performance issues between admin/non-admin | 13:31 |
*** sdake has joined #openstack-infra | 13:31 | |
rcarrillocruz | that change is about putting on the server dict the fips manually | 13:32 |
odyssey4me | pabelanger I'm not sure I understand where it is that tox uses the job name? | 13:32 |
fungi | pabelanger: odyssey4me: fun. i wonder if we should just switch to a hash of job names for our workspaces | 13:32 |
pabelanger | fungi: ya, that would solve it | 13:32 |
rcarrillocruz | since different clousd may populate neutron -> nova fips info slower thatn others | 13:32 |
*** jpich has quit IRC | 13:32 | |
rcarrillocruz | i've noticed the timeouts on devstack instances, whre it didn't before | 13:32 |
fungi | odyssey4me: i think it's that we end up with /home/jenkins/workspace/some-really-really-long-job-name/.tox/some-fairly-long-env-name/bin/python and so on | 13:32 |
rcarrillocruz | and that same issue was also hit by asselin__ , but he uses a HOS cloud | 13:32 |
*** d0ugal has quit IRC | 13:33 | |
odyssey4me | fungi ah, so the workspace is based on the job name - not the repo or something else | 13:33 |
*** jtomasek_ has quit IRC | 13:33 | |
rcarrillocruz | long story short: the fix would give us fast create_server with auto_ip, which would fix the test issues, but regardless we should investigate at some point why the calls are so slow for non-admin users | 13:33 |
pabelanger | rcarrillocruz: mind reviewing 340937 | 13:33 |
*** _ari_ has joined #openstack-infra | 13:33 | |
rcarrillocruz | sure | 13:33 |
*** ajmiller has quit IRC | 13:33 | |
*** ajmiller has joined #openstack-infra | 13:33 | |
pabelanger | fungi: ^ if you are also online too | 13:33 |
*** sfinucan has quit IRC | 13:34 | |
*** baoli_ has quit IRC | 13:34 | |
fungi | odyssey4me: right. honestly that's just there to accommodate workers which persist and run multiple different jobs. for our single-use workers we could just use a generic workspace because they'll never get reused | 13:34 |
rcarrillocruz | lgtm | 13:34 |
fungi | odyssey4me: however that's not configurable at the moment | 13:34 |
openstackgerrit | Alex Kavanagh proposed openstack-infra/project-config: Add charm-barbican, and related https://review.openstack.org/340945 | 13:34 |
fungi | odyssey4me: we're just mirroring jenkins' old behavior | 13:35 |
odyssey4me | fungi hmm, so using a hash of the job name or a mktemp folder could work | 13:35 |
*** d0ugal_ is now known as d0ugal | 13:35 | |
*** mtanino has joined #openstack-infra | 13:36 | |
fungi | odyssey4me: right | 13:36 |
odyssey4me | now that jenkins is no more, things can change | 13:36 |
*** d0ugal is now known as Guest65715 | 13:36 | |
*** sfinucan has joined #openstack-infra | 13:36 | |
*** Guest65715 has quit IRC | 13:36 | |
*** d0ugal_ has joined #openstack-infra | 13:36 | |
fungi | well, probably not mktemp on our long-lived nodes because we may want to reuse the same workspace for multiple runs of the same job, but a deterministic hash would be fine | 13:37 |
*** baoli has joined #openstack-infra | 13:37 | |
fungi | as long as the collision rate is reasonably infinitesimal | 13:37 |
*** baoli_ has joined #openstack-infra | 13:39 | |
odyssey4me | fungi or perhaps just let the workspace name be optionally set to something static or something random | 13:39 |
*** ddieterly is now known as ddieterly[away] | 13:39 | |
*** Nikolay_St has joined #openstack-infra | 13:39 | |
fungi | yep | 13:39 |
*** jaosorior has quit IRC | 13:40 | |
*** Na3iL has quit IRC | 13:40 | |
odyssey4me | where does the workspace get defined? | 13:40 |
pabelanger | zuul-launcher | 13:40 |
odyssey4me | let me see if I can figure something out | 13:41 |
fungi | anyway, problem now is that we really don't want to extend zuul-launcher. we considered feature parity with the parts of jenkins we used sufficient to make it feature-complete and work has since shifted back to zuul v3 which is pretty drastically different | 13:41 |
*** sdake_ has joined #openstack-infra | 13:41 | |
*** caowei has joined #openstack-infra | 13:41 | |
*** ramishra has quit IRC | 13:41 | |
fungi | so we need a pretty strong case for altering the "old" zuul-launcher codebase at this point | 13:41 |
*** eharney has joined #openstack-infra | 13:42 | |
*** baoli has quit IRC | 13:42 | |
*** mtanino has quit IRC | 13:43 | |
*** ramishra has joined #openstack-infra | 13:43 | |
fungi | as zuul-launcher was merely intended as a stop-gap to get us off jenkins while zuul v3 is still under development, and it isn't really expected to have a future once v3 is production ready (though some ideas from it may be reused in v3) | 13:43 |
*** jamesmcarthur has joined #openstack-infra | 13:43 | |
*** notnownikki has joined #openstack-infra | 13:43 | |
*** d0ugal_ has quit IRC | 13:44 | |
odyssey4me | fungi hmm, then perhaps you can help me work out a way to achieve the goal I have without extending zuul-launcher | 13:44 |
*** d0ugal has joined #openstack-infra | 13:44 | |
odyssey4me | fungi with reference to https://review.openstack.org/339770 | 13:45 |
*** amotoki has joined #openstack-infra | 13:45 | |
*** mdrabe has quit IRC | 13:46 | |
openstackgerrit | Kirill Zaitsev proposed openstack-infra/project-config: Add {name}-python27-mitaka jobs to murano projects https://review.openstack.org/340954 | 13:47 |
*** matthewbodkin has joined #openstack-infra | 13:47 | |
fungi | Zara: odyssey4me the general way to do that is to use separate job parameters for the name substring and whatever you actually want set in the job, so that you can still make the jobs have unique names but are able to pass longer strings into the job payload | 13:48 |
*** ddieterly[away] is now known as ddieterly | 13:48 | |
fungi | er, sorry Zara. tab completion failure | 13:48 |
fungi | odyssey4me: ^ | 13:48 |
Zara | haha | 13:48 |
*** ramishra_ has joined #openstack-infra | 13:48 | |
Zara | I was worried for a second xD | 13:48 |
*** e0ne has quit IRC | 13:49 | |
*** Goneri has quit IRC | 13:49 | |
*** ramishra has quit IRC | 13:49 | |
openstackgerrit | Merged openstack-infra/project-config: Fix URL variable with reprepro-import builder https://review.openstack.org/340937 | 13:50 |
*** PollyZ has joined #openstack-infra | 13:50 | |
odyssey4me | fungi in this case the only thing I need to protect is the translation of 'func' to the tox env 'functional' - this is to prevent having to patch many, many repositories with a tox config change... | 13:51 |
*** amrith has quit IRC | 13:51 | |
*** amrith has joined #openstack-infra | 13:52 | |
odyssey4me | fungi but I would like to allow the optional use of an alternative tox env, without building a new and almost identical job template | 13:52 |
*** e0ne has joined #openstack-infra | 13:53 | |
*** ramishra_ has quit IRC | 13:53 | |
*** rbrndt has joined #openstack-infra | 13:53 | |
*** garyk has quit IRC | 13:53 | |
*** r-mibu has quit IRC | 13:54 | |
*** r-mibu has joined #openstack-infra | 13:54 | |
*** ramishra has joined #openstack-infra | 13:55 | |
*** ajmiller has quit IRC | 13:55 | |
*** ajmiller has joined #openstack-infra | 13:55 | |
*** gongysh has joined #openstack-infra | 13:56 | |
*** caowei has quit IRC | 13:57 | |
openstackgerrit | Merged openstack-infra/tripleo-ci: Use venv for all tempest commands https://review.openstack.org/339118 | 13:57 |
*** Goneri has joined #openstack-infra | 13:57 | |
openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard-webclient: Add quick non-permanent filter buttons for timeline events https://review.openstack.org/331748 | 13:59 |
*** xarses_ has quit IRC | 13:59 | |
*** yamamoto has quit IRC | 13:59 | |
waynr | electrofelix: i haven't had a chance to review them yet | 13:59 |
fungi | you could adjust the template so that the env you pass to the tox builder is a separate parameter from the name substring you insert, and then pass both when instantiating the template in a project | 13:59 |
fungi | odyssey4me: ^ | 13:59 |
*** Julien-zte has joined #openstack-infra | 13:59 | |
*** sarob has joined #openstack-infra | 13:59 | |
*** isaacb has quit IRC | 14:00 | |
*** sarob has quit IRC | 14:00 | |
*** sarob has joined #openstack-infra | 14:00 | |
fungi | odyssey4me: pabelanger: for reference, i believe this is the issue we're attempting to work around https://github.com/pypa/virtualenv/issues/596 | 14:01 |
fungi | or https://github.com/pypa/pip/issues/1773 | 14:02 |
fungi | though as those issue discussions point out, it's more of a general shortcoming of shebang lines on linux, so anything we do is going to simply be working around that | 14:03 |
*** burgerk has quit IRC | 14:03 | |
odyssey4me | yeah, so it seems | 14:03 |
*** rdx_matt has joined #openstack-infra | 14:04 | |
*** piet has quit IRC | 14:04 | |
odyssey4me | I suppose in the jobs it could be easy enough to create a mktemp folder name, symlink that name to the workspace directory, then execute tox from inside the symlinked folder | 14:05 |
odyssey4me | but that's a horrible hack | 14:05 |
*** javeriak has quit IRC | 14:05 | |
fungi | yeah, i still definitely think this is something we should fix by not using job names in workspaces with zuul, it's just a question of whether that's going to happen in zuul v2.5's zuul-launcher, or in zuul v3, or both | 14:06 |
odyssey4me | sure, unfortunately with that option there will needs to be some discussion and it'll take more time than I can afford... so while I'm happy to work on that as a mid to long term option, I need something for the short term that will work around the limitation | 14:08 |
fungi | the end result of the present design is that the job name and the tox env name added together can't exceed something like 86 characters | 14:08 |
pabelanger | fungi: good find | 14:09 |
odyssey4me | yeah, and we have a particularly long repo name which resulted in the folder name of /home/jenkins/workspace/gate-openstack-ansible-lxc_container_create-ansible-functional-ubuntu-xenial-nv/.tox/functional/bin/python | 14:09 |
odyssey4me | that's 131 chars | 14:09 |
*** berendt has quit IRC | 14:09 | |
odyssey4me | thus we changed 'functional' to 'func' in the job name | 14:10 |
fungi | so anyway, separating the job name uniqueness parameter from the tox env parameter is one way to work around it | 14:10 |
openstackgerrit | Ihar Hrachyshka proposed openstack-infra/project-config: Clone grenade for multinode linuxbridge job https://review.openstack.org/340962 | 14:10 |
*** markvoelker has quit IRC | 14:11 | |
fungi | that way you can set strings per instantiation of the template to make the resulting job names unique without tying them to the length of the tox env names | 14:11 |
odyssey4me | is there a way that I can do that without having to specify the list of valid options in the job or job group? | 14:12 |
fungi | odyssey4me: not sure what you mean by "the list of valid options" | 14:12 |
*** hurgleburgler has joined #openstack-infra | 14:13 | |
fungi | odyssey4me: you would just pass multiple parameters to each instantiation of the job-template (whether that's in a job-group or in a project) | 14:13 |
*** gordc has joined #openstack-infra | 14:13 | |
tlbr | pabelanger, packages for packaging-deb project are under https://tarballs.openstack.org/packaging-deb/*/uploads/*/ path. It was discussed with mordred and fungi on #openstack-pkg channel | 14:13 |
*** zz_dimtruck is now known as dimtruck | 14:13 | |
*** wznoinsk_ has joined #openstack-infra | 14:14 | |
*** timburke_ has joined #openstack-infra | 14:15 | |
pabelanger | tlbr: that is an easy fix, however I'd like want to drop the packaging-deb portion of the URL, so we can make our JJB build generic for other project to consume, if they wanted too | 14:15 |
*** alvladeru has joined #openstack-infra | 14:15 | |
*** timburke has quit IRC | 14:15 | |
*** wznoinsk has quit IRC | 14:15 | |
*** Nakato has quit IRC | 14:15 | |
*** aderyugin has quit IRC | 14:15 | |
*** jlk has quit IRC | 14:15 | |
*** karthikp_ has joined #openstack-infra | 14:15 | |
*** timrc has quit IRC | 14:15 | |
*** sigmavirus has quit IRC | 14:15 | |
*** alvladeru is now known as aderyugin | 14:15 | |
odyssey4me | fungi hmm, I think the two wires have connected and I have seen the error of my ways... let me revise that patch | 14:15 |
*** jlk` has joined #openstack-infra | 14:15 | |
*** links has quit IRC | 14:15 | |
*** timrc has joined #openstack-infra | 14:16 | |
*** Nakato has joined #openstack-infra | 14:16 | |
fungi | pabelanger: the problem there is that they want the parent directory of, for example, the nova packages to just be called "nova" but want to avoid them ending up in the same place as the nova project's tarballs | 14:17 |
*** yanyanhu has quit IRC | 14:17 | |
fungi | pabelanger: which is why they prefixed them with that packaging-deb directory | 14:17 |
*** tonytan_brb has joined #openstack-infra | 14:17 | |
zigo | pabelanger: I'd prefer not to. I'll do the packaging under packaging-deb for these stuff we want to package for Debian / Ubuntu as well. | 14:17 |
fungi | i missed that 340937 was removing that | 14:17 |
*** timburke_ is now known as timburke | 14:18 | |
pabelanger | fungi: okay, that's fine. Like I said, an easy fix | 14:18 |
pabelanger | once we move it into post pipeline we can add the prefix | 14:18 |
*** sputnik13 has quit IRC | 14:18 | |
openstackgerrit | Spyros Trigazis proposed openstack-infra/project-config: [magnum] Add ironic projects to PROJECTS env https://review.openstack.org/340965 | 14:18 |
*** Julien-zte has quit IRC | 14:18 | |
*** Julien-zte has joined #openstack-infra | 14:19 | |
zigo | pabelanger: Thanks for this work, btw, I was getting depressed seeing no progress. | 14:19 |
fungi | but to pabelanger's point, people should also be free to reuse these same mechanisms to create deb packages of repos outside the control of the packaging-deb team, so will end up either creating a separate and nearly-identical job with a different upload namespace or will need the existing template to be slightly more generic | 14:19 |
electrofelix | regarding https://review.openstack.org/320994 a change for elastic-recheck, although it was abandoned as superseded, a question remained unanswered, that is why is '-U' being used for dependency installation? | 14:19 |
*** dingyichen has quit IRC | 14:20 | |
*** tonytan4ever has quit IRC | 14:20 | |
*** sigmavirus has joined #openstack-infra | 14:20 | |
*** ramishra has quit IRC | 14:20 | |
pabelanger | infra-root: any objections on helping landing 334652? I'd like to enable SSH pipelining for ansible on puppetmaster.o.o, that should result in a performance increase and help reduce our wheel under 15mins again | 14:20 |
*** mdrabe has joined #openstack-infra | 14:20 | |
electrofelix | it does appear to prevent upper constraints from being adhered to when dealing with test-requirements | 14:20 |
zigo | fungi: In such case, we can call the destination folder something else than "packaging-deb"... | 14:20 |
zigo | The only point is to use a generic name for the dest folder for the pickup job to read from. | 14:21 |
fungi | zigo: right, the question is whether the existing job-template can be made more generic to accommodate that, or whether we have a separate job-template for each | 14:21 |
openstackgerrit | Merged openstack-infra/storyboard-webclient: Add quick non-permanent filter buttons for timeline events https://review.openstack.org/331748 | 14:21 |
zigo | I don't intend to slow down anyone by doing more bikeshedding ! :) | 14:21 |
pabelanger | fungi: zigo: Ya, the naming to me is moot honestly, its a place to store tmp files. We should consider something along that lines which gets purged every x days | 14:21 |
zigo | As you like, as long as we get it done. | 14:21 |
fungi | but i also think that until such time as there are people who want to do that (build debs and upload them somewhere other than packaging-deb) we can live with that hard-coded | 14:22 |
pabelanger | zigo: yes, lets get something working first, then iterate on that | 14:22 |
*** karthikp_ has quit IRC | 14:22 | |
fungi | pabelanger: is that pipelining change what you were discussing with jeblair, Shrews, et al last night? | 14:23 |
*** acoles_ is now known as acoles | 14:23 | |
*** ramishra has joined #openstack-infra | 14:24 | |
*** EricGonczer_ has joined #openstack-infra | 14:27 | |
*** chem|away has joined #openstack-infra | 14:27 | |
*** mriedem1 has joined #openstack-infra | 14:28 | |
electrofelix | mtreinish: would you mind responding to the follow up question on https://review.openstack.org/320994 ? | 14:28 |
electrofelix | mtreinish: It seems to me that the way tox works by performing using '{install_command} -e .' at the end when usedevelop = True, if you put '-U' into the install_command anything installed via test-requirements can be upgraded past it's upper pins | 14:28 |
*** chem|away is now known as chem | 14:28 | |
electrofelix | mtreinish: this seems to only occur with test-requirements.txt since pbr reflects requirements.txt into setuptools so upper pins will be adhered to (assuming correct ordering) by a the install_command. | 14:28 |
*** piet has joined #openstack-infra | 14:28 | |
electrofelix | mtreinish: so I'm wondering whether '-U' is inherently problematic for test-requirements.txt and upper constraints | 14:29 |
*** _oanson has joined #openstack-infra | 14:29 | |
openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Improve logger output for expanding templates https://review.openstack.org/336091 | 14:29 |
openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Refactor base test classes inheritance for reuse https://review.openstack.org/336090 | 14:29 |
*** mriedem has quit IRC | 14:29 | |
clarkb | pabelanger: is the sudo requiretty thing a non issue? we ssh as root for puppet things but zuul launcher will ssh as jenkins/zuul | 14:30 |
clarkb | also I wish the docs explained what this did | 14:30 |
*** markvoelker has joined #openstack-infra | 14:30 | |
*** isaacb has joined #openstack-infra | 14:31 | |
*** julim has joined #openstack-infra | 14:31 | |
openstackgerrit | Darragh Bailey proposed openstack-infra/elastic-recheck: SearchEngine search only supports ints for days https://review.openstack.org/320988 | 14:31 |
*** mriedem1 is now known as mriedem | 14:31 | |
pabelanger | fungi: no, this is for our ansible / puppet wheel. That was related to zuul-launcher | 14:31 |
clarkb | electrofelix: pip will respect the upper bounds in the test requirements file too | 14:31 |
clarkb | I mean as much as it respects them at all | 14:32 |
*** eharney has quit IRC | 14:32 | |
*** mtanino has joined #openstack-infra | 14:32 | |
electrofelix | clarkb: it doesn't appear to when usedevelop=True | 14:32 |
pabelanger | clarkb: I don't believe so, however I have not tested it. When I looked for zuul-launcher, we were not affected. However, until we test with puppetmaster and puppet, I am not 100% sure | 14:32 |
*** amitgandhinz has quit IRC | 14:33 | |
electrofelix | clarkb: or rather when usedevelop=True and -U is in the install command | 14:33 |
*** gongysh has quit IRC | 14:33 | |
pabelanger | clarkb: I _think_ the issue is if you call becomes: directly from your ansible-playbook | 14:33 |
electrofelix | so I left an open question on a review as to whether it was a good idea to have -U enabled | 14:33 |
clarkb | electrofelix: the reason it is there is so you dont have to tox -r | 14:33 |
*** ajmiller has quit IRC | 14:33 | |
*** amitgandhinz has joined #openstack-infra | 14:33 | |
pabelanger | that will requiretty settings | 14:34 |
clarkb | and -U is very intentional amd shpuld not do ehat you describe | 14:34 |
*** ajmiller has joined #openstack-infra | 14:34 | |
clarkb | -U means upgradw within the bpunds provoded not install newest everything unconditionally | 14:34 |
*** abregman is now known as abregman|mtg | 14:34 | |
electrofelix | I believe the problem is that test-requirements.txt is not included in the final 'pip install -e .' when usedevelop=True | 14:34 |
clarkb | pabelanger: ya which we do for the zuul laumcher restart playbook but nowhere else that I know of | 14:34 |
electrofelix | requirements.txt is reflected into it via setuptools | 14:35 |
pabelanger | clarkb: Right, that's the only place I can think of it too | 14:35 |
electrofelix | which means -U + usedevelop=True does not respect upper constraints in test-requirements.txt, but would for requirements.txt | 14:35 |
electrofelix | arguably that's a bug in tox | 14:35 |
clarkb | electrofelix: yes but at that point both requirements and testrequiremts have been preinstall and the install of . should just do . | 14:36 |
*** ajmiller has quit IRC | 14:36 | |
*** ajmiller has joined #openstack-infra | 14:36 | |
*** tonytan4ever has joined #openstack-infra | 14:36 | |
openstackgerrit | Merged openstack-infra/tripleo-ci: Switch away from the default IP range used by tripleo https://review.openstack.org/338886 | 14:36 |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 14:36 |
*** garyk has joined #openstack-infra | 14:37 | |
*** mctaylor has joined #openstack-infra | 14:38 | |
*** karthikp_ has joined #openstack-infra | 14:38 | |
*** mctaylor has quit IRC | 14:38 | |
*** rdx_matt has quit IRC | 14:38 | |
electrofelix | clarkb: I don't think test-requirements is being passed to the final "pip install -U -e ." so the upper bound is not defined | 14:38 |
*** tonytan_brb has quit IRC | 14:39 | |
*** rossella_s has quit IRC | 14:40 | |
clarkb | electrofelix: is the issue where they overlap but you arent apecific enough in requirements? | 14:40 |
clarkb | electrofelix: pip shpuldnt look at everything installed amd upgrade it, it doesnt work like apt that way | 14:40 |
*** Guest94106 has quit IRC | 14:40 | |
clarkb | it will only touch things in the input | 14:40 |
*** Hal has joined #openstack-infra | 14:40 | |
*** rossella_s has joined #openstack-infra | 14:40 | |
clarkb | so test libs like fixtures are left as is because they are not in the requirements.txt set | 14:41 |
*** Hal is now known as Guest19946 | 14:41 | |
*** sree has quit IRC | 14:41 | |
electrofelix | I'm concerned that it's behaviour doesn't match what was stated as to how it should be behaving | 14:41 |
*** sorantis has quit IRC | 14:42 | |
*** clenimar has joined #openstack-infra | 14:43 | |
*** clenimar_ has joined #openstack-infra | 14:43 | |
*** clenimar_ has quit IRC | 14:43 | |
*** clenimar has quit IRC | 14:43 | |
openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard-webclient: Allow comments to be edited by their owner https://review.openstack.org/333418 | 14:43 |
*** clenimar has joined #openstack-infra | 14:44 | |
*** deadnull has joined #openstack-infra | 14:44 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 14:44 |
fungi | electrofelix: it may be easier to provide a short reproducer for the observed (mis)behavior | 14:45 |
fungi | to make it easier to rule out other more likely causes | 14:45 |
*** david-lyle has joined #openstack-infra | 14:46 | |
electrofelix | fungi clarkb: so http://paste.openstack.org/show/530896/ shows the exact behaviour using pip commands, I debugged tox before to show me what was being executed. testtools has a lower pin on fixtures, but pip seems to ignore this in favour of the newer version | 14:48 |
electrofelix | sorry, meant pip ignores the fact that the installed version is sufficient and still upgrades it to a newer version | 14:48 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Only run kill if there is a httpserver to kill https://review.openstack.org/340829 | 14:48 |
openstackgerrit | Kirill Zaitsev proposed openstack-infra/project-config: Add {name}-python27-mitaka jobs to murano projects https://review.openstack.org/340954 | 14:49 |
clarkb | electrofelix: roght so the bug there is not having fixtures in requirements.txt | 14:49 |
electrofelix | fixtures however is in test-requirements.txt with an upper constraint | 14:49 |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 14:49 |
*** eezhova has quit IRC | 14:49 | |
fungi | electrofelix: just to be clear, when you say "upper constraint" you mean in a constraint list specified with the -c option to pip? | 14:49 |
fungi | electrofelix: or just "an upper bound in a version specifier"? | 14:50 |
electrofelix | test-requirements.txt contains a line with 'fixtures>=0.3.14,<3.0' | 14:50 |
fungi | electrofelix: aha, communication breakdown. that's not a "constraint" | 14:50 |
clarkb | electrofelix: yes you need to either update that bound to be valid or move fixtures into requirements.txt | 14:50 |
openstackgerrit | Steve Martinelli proposed openstack-infra/project-config: publish osc-lib docs https://review.openstack.org/340981 | 14:50 |
electrofelix | sorry reusing a word with different meaning? | 14:50 |
electrofelix | clarkb: how is that not valid? | 14:51 |
*** mariusv has joined #openstack-infra | 14:51 | |
clarkb | electrofelix: pip has "constraints" which are install this version despite any other bounds listings | 14:51 |
fungi | electrofelix: pip does not have a holistic dep solver, it evaluates each version specifier as it encounters it. this is a very long-standing implementation shortcoming in pip which constraints (-c) is a workaround for | 14:51 |
clarkb | electrofelix: because of how pip works its required... pip is not apt and will not remember you had a previous bound | 14:51 |
*** zeih has joined #openstack-infra | 14:51 | |
clarkb | ya what fungi said basically | 14:51 |
electrofelix | so basically you're agreeing there is a hole between test-requirements.txt use of -U and upper bounds | 14:52 |
clarkb | no | 14:52 |
*** xarses has joined #openstack-infra | 14:52 | |
fungi | electrofelix: it's pip working as designed | 14:52 |
fungi | electrofelix: the design is simply not so great if you expect behavior like a distro package manager | 14:52 |
*** xarses has quit IRC | 14:52 | |
clarkb | right pip is doing what you told it. You can achieve the behavior you want by telling pip something else | 14:52 |
electrofelix | to be clear, testtools is not specified in requirements either | 14:53 |
*** xarses has joined #openstack-infra | 14:53 | |
electrofelix | it only appears in test-requirements.txt | 14:53 |
*** gongysh has joined #openstack-infra | 14:53 | |
electrofelix | so basically any upper bound in test-requirements.txt is going to get ignored | 14:53 |
electrofelix | once usedevelop=True and -U is set on the pip install command | 14:53 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Remove the rewrite of TOCI_JOBTYPE https://review.openstack.org/338566 | 14:53 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 14:53 |
*** nmagnezi has quit IRC | 14:54 | |
*** hurgleburgler has quit IRC | 14:54 | |
*** burgerk has joined #openstack-infra | 14:54 | |
fungi | electrofelix: one very fiddly workaround is to reorder your requirements and test requirements to make sure the most specific version specifiers of deps (including transitive deps in a depth-first walk if memory serves) are encountered first | 14:54 |
electrofelix | AFAICT the only reason upper bound works in requirements.txt is because pbr will reflect those into setuptools and result in pip seeing them even if tox does not pass '-r requirements.txt' | 14:54 |
*** jed56 has quit IRC | 14:55 | |
electrofelix | fungi: this doesn't work no matter when I list the fixtures and upper bound in test-requirements.txt | 14:55 |
clarkb | electrofelix: no in fact if it reflected nothing you would get the behavior you eant | 14:55 |
clarkb | b3cause pip unlike apt only works on its current inputs | 14:55 |
fungi | electrofelix: we relied very, very heavily on "lucky ordering" of dependencies in our requirements lists for a long time before constraints was implemented. now we use constraints lists to set specific solved sets of dependencies alongside the requirements lists | 14:56 |
*** rockstar_ is now known as rockstar | 14:57 | |
openstackgerrit | Sagi Shnaidman proposed openstack-infra/gear: Support TCP keepalives in geard https://review.openstack.org/334452 | 14:57 |
fungi | electrofelix: there's a lot of explanation in http://git.openstack.org/cgit/openstack/requirements/tree/README.rst if you're ready to be extremely disappointed by the state of pip's solver | 14:57 |
openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 14:57 |
electrofelix | clarkb: so if pip was only getting it's current inputs, how is it getting the dependency on testtools to process? | 14:58 |
fungi | electrofelix: i haven't looked at your paste closely yet, but odds are it's a transitive dependency of one of the specified deps | 14:58 |
electrofelix | fungi: so it's not really a test only dep? | 14:59 |
fungi | Collecting fixtures>=1.3.0 (from testtools>=0.9.34->python-subunit>=0.0.18->subunit2sql>=0.9.0->elastic-recheck==0.0.1.dev1770) | 14:59 |
*** cody-somerville has joined #openstack-infra | 14:59 | |
*** yamamoto has joined #openstack-infra | 14:59 | |
electrofelix | yeap just spotted that as well | 15:00 |
electrofelix | so it's not really a test dependency at all | 15:00 |
fungi | electrofelix: so elastic-recheck wants a version of subunit2sql which wants a version of python-subunit which wants a version of testtools which wants fixtures>=1.3.0 (no upper bound on the specifier) | 15:00 |
electrofelix | so should be moved to requirements.txt to prevent surprises in the future? | 15:00 |
fungi | electrofelix: that's one workaround, yes | 15:01 |
fungi | the other is to make a constraints file containing an == to a specific version of fixtures you know meets all the needs of everything else in your dep chain and pass that with -c every time pip is called | 15:01 |
fungi | which is how it's being handled in most openstack projects now | 15:01 |
*** jamesmca_ has joined #openstack-infra | 15:02 | |
*** jamesmcarthur has quit IRC | 15:02 | |
*** yaume_ has joined #openstack-infra | 15:02 | |
electrofelix | sounds a bit like how pip-compile was pitched | 15:03 |
*** bin_ has joined #openstack-infra | 15:03 | |
*** weshay_mtg is now known as weshay | 15:05 | |
*** matbu is now known as matbu|mtg | 15:05 | |
*** yaume has quit IRC | 15:05 | |
fungi | electrofelix: yes, we do a process similar to pip-compile to generate our upper-constraints.txt from the state of global-requirements.txt\ | 15:05 |
electrofelix | fungi clarkb: thanks with being so patient with me, this stuff was driving me around the bend | 15:06 |
fungi | and then we run some canary tests with the proposed upper-constraints.txt before merging, and then other projects pull upper-constraints.txt and apply it in their tox invocations | 15:06 |
*** ihrachys has quit IRC | 15:06 | |
*** yamamoto has quit IRC | 15:07 | |
*** samueldmq has joined #openstack-infra | 15:07 | |
*** nadya has quit IRC | 15:07 | |
*** cardeois_ has joined #openstack-infra | 15:07 | |
*** hashar is now known as hasharAway | 15:07 | |
samueldmq | hi folks | 15:07 |
electrofelix | fungi: looks like this could do to be added to the elastic-recheck, I'll no | 15:07 |
samueldmq | I have a question about bindep/other-requirements | 15:07 |
electrofelix | I'll have a look in a few days at submitting it | 15:07 |
clarkb | samueldmq: ask away | 15:08 |
*** hurgleburgler has joined #openstack-infra | 15:08 | |
openstackgerrit | Paul Belanger proposed openstack-infra/project-config: Enable reprepro-release-deb-openstack https://review.openstack.org/340993 | 15:08 |
*** Ravikiran_K has quit IRC | 15:08 | |
openstackgerrit | Merged openstack-infra/tripleo-ci: Upload images to cache at end of ovb job https://review.openstack.org/338762 | 15:08 |
samueldmq | clarkb: other-requirements is expected to list deps for a dev environment ? | 15:09 |
pabelanger | fungi: clarkb: ^ mind a review? reprepro import is working so I've enabled reprepro-release to test our AFS release process | 15:09 |
samueldmq | clarkb: or for a production/deployment environment? | 15:09 |
pabelanger | adds some new mutexes | 15:09 |
*** hurgleburgler1 has joined #openstack-infra | 15:10 | |
*** ramishra_ has joined #openstack-infra | 15:10 | |
clarkb | samueldmq: it depends on who you ask ... I have asserted it should be for dev work because that is the most likely target installing from source. But I think we have decided to have a [test] env and a default env for production like deps | 15:10 |
*** cardeois has quit IRC | 15:10 | |
*** yamahata has joined #openstack-infra | 15:10 | |
clarkb | no andreas here but fungi ^ is that how we are doing it? | 15:11 |
*** isaacb has quit IRC | 15:11 | |
samueldmq | zigo: ^ | 15:12 |
fungi | samueldmq: it's flexible. any packages you list without a user profile or with a "test" user profile will be identified if the platform profiles match your target environment | 15:12 |
fungi | samueldmq: so you _can_ use [test] to indicate it's a test-only dependency, and not put [test] on runtime deps | 15:12 |
fungi | samueldmq: or you can just ignore it and only bother listing your test deps | 15:13 |
*** hurgleburgler has quit IRC | 15:13 | |
*** ramishra has quit IRC | 15:13 | |
fungi | our jobs will support either method | 15:13 |
zigo | fungi: From my perspective (ie: packaging), how should I read it? | 15:13 |
pabelanger | also note, we have no way to check test-only or runtime deps currently in bindep jobs | 15:13 |
*** dronshaw has joined #openstack-infra | 15:13 | |
zigo | fungi: No [test] means it's also runtime ? | 15:13 |
fungi | zigo: i don't know that it necessarily feeds packager decisions | 15:14 |
*** ams__ has quit IRC | 15:14 | |
samueldmq | fungi: clarkb: for now we only have a list in keystone's file: https://github.com/openstack/keystone/blob/master/other-requirements.txt | 15:14 |
*** notnownikki has quit IRC | 15:14 | |
zigo | fungi: For the case of keystoneclient, I see real mistakes though. | 15:14 |
fungi | zigo: no [test] _might_ mean it's a runtime dep, or might mean that the project is only using that list to specify their test deps | 15:14 |
zigo | :/ | 15:15 |
samueldmq | does that mean they're all runtime deps ? | 15:15 |
clarkb | my xenial test loop is at project 268 of 305 | 15:15 |
clarkb | of course all the db projects are still to run and will likely be slower :( | 15:16 |
pabelanger | clarkb: nice, anything explode yet? | 15:16 |
fungi | zigo: the main use case we've targeted is, as a developer i clone python-keystoneclient and want to test some changes i've made by running `tox`, but then a bunch of the python dependencies which it tries to pip install won't compile because i'm missing c headers. with a sane other-requirements.txt file in that repo, running `bindep` (or maybe `tox -e bindep`) will tell me what distro packages i need | 15:16 |
fungi | to install first | 15:16 |
clarkb | samueldmq: fungi is saying that if a project does not have explicit test deps in that list annotated with [test] then they may or not be test/dev deps | 15:16 |
clarkb | pabelanger: pbr is probably the biggest worry I have right now since its pretty central | 15:16 |
clarkb | pabelanger: its wsgi tests were unhappy for some reason, I will ahve to do more digging after this initial pass through | 15:17 |
clarkb | pabelanger: otherwise from what I have seen its either tests are failign on trusty too or they have no tests or code yet | 15:17 |
zigo | fungi: How come we get stuff like python3-dev in [platform:dpkg] ? | 15:17 |
zigo | fungi: There's no need for Python.h in ksc ... | 15:17 |
openstackgerrit | Spyros Trigazis proposed openstack-infra/project-config: [magnum] Add ironic projects to PROJECTS env https://review.openstack.org/340965 | 15:18 |
*** adrian_otto has joined #openstack-infra | 15:18 | |
clarkb | zigo: because ksc likely builds some dep on source installs that links against python | 15:18 |
*** ihrachys has joined #openstack-infra | 15:18 | |
samueldmq | clarkb: nice. so for keystone I understand it as setting up a _dev_ env, as it reflects http://docs.openstack.org/developer/keystone/setup.html#installing-dependencies | 15:18 |
clarkb | lxml probably | 15:18 |
fungi | zigo: are you able to pip install all of ksc's requirements.txt and test-requirements.txt without that installed first? | 15:18 |
samueldmq | clarkb: I am going to update those docs do advise using bindep | 15:18 |
*** sarob has quit IRC | 15:19 | |
zigo | fungi: Just by a quick look, I'd say yes. | 15:19 |
*** acoles is now known as acoles_ | 15:19 | |
fungi | anyway, we made our jobs look for the test user profile (in addition to the implicit default profile) at the request of some projects who wanted to get more specific about what packages are runtime non-python deps vs which ones are test-time non-python deps, but it's not expected to be consistent between projects | 15:19 |
fungi | zigo: including in a python 3 virtualenv? | 15:19 |
zigo | No idea about venv, but if that's the case we need it for venv, then it will appear everywhere, which isn't probably what we want, is it? | 15:20 |
fungi | zigo: easiest way to test that theory is to remove that line from the other-requirements.txt file and propose that change to the python-keystoneclient repo and see what jobs fail (if any) | 15:20 |
samueldmq | fungi: zigo: and that will also depends on the distribution, some may meet all the requirements, others may meet none of them | 15:20 |
fungi | zigo: it will only appear for projects which need to compile non-python extensions. that's certainly not _everywhere_ | 15:20 |
samueldmq | having a list of deps and using bindep will ensure consistency | 15:20 |
clarkb | pabelanger: http://paste.openstack.org/show/530926/ thats the current list of results | 15:20 |
clarkb | pabelanger: note that in most cases releasenotes failing is due to no releasenotes tox target which is fine | 15:21 |
apuimedo | mordred: the condorcet voting system tells me my voter key is invalid | 15:21 |
clarkb | dhellmann: ^ you might also be curious for the current data | 15:21 |
zigo | fungi: I also saw stuff like "python3.4 [platform:ubuntu-trusty]", that's completely wrong, it should *always* be replaced by python3-all (which means: all available python in the system) | 15:21 |
clarkb | apuimedo: there is a mailing list thread about it. mordred is trying to sort out what happened | 15:21 |
apuimedo | ah, ok | 15:21 |
apuimedo | sorry I didn't see it | 15:21 |
fungi | apuimedo: see his followup to the -dev ml. there was something wrong with the ballots. he's going to try again, not sure when | 15:21 |
apuimedo | thanks :-) | 15:21 |
apuimedo | you guys are always on top of things :-) | 15:22 |
*** tidwellr has joined #openstack-infra | 15:22 | |
clarkb | pabelanger: looks like osc failed py27 I should check on that | 15:23 |
*** andymaier has quit IRC | 15:23 | |
dhellmann | clarkb : lots of release notes and doc build failures. I wonder how many of those are related to the issue fixed in https://review.openstack.org/#/c/334201/ | 15:23 |
*** cardeois_ has quit IRC | 15:23 | |
fungi | zigo: sounds reasonable. i compiled the fallback list by looking at what packages we were installing from puppet on our old workers where we used to preinstall a lot of this stuff, and then entries from that have been getting copied to other-requirements.txt files in individual projects. i likely missed some cases where we'd have been better to install a virtual package rather than whatever it refers | 15:24 |
fungi | to | 15:24 |
*** cardeois_ has joined #openstack-infra | 15:24 | |
clarkb | dhellmann: at least for releasenotes I think most of them are my script trying to run that tox target when it doesn't exist | 15:24 |
zigo | fungi: I'll try and fix what I can manually. | 15:24 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 15:24 |
dhellmann | clarkb : ah | 15:24 |
clarkb | dhellmann: I will run thatthrough a filter to remove the fails caused by that when I get closer to producing final results | 15:24 |
*** david-lyle has quit IRC | 15:24 | |
dhellmann | sounds good | 15:24 |
clarkb | AttributeError: 'module' object has no attribute 'select_auth_plugin' is the osc issue | 15:25 |
fungi | zigo: all of this can be fixed by just proposing corrections to the relevant files. also http://codesearch.openstack.org/ may be useful in looking for which files in which projects have those known bad entries | 15:25 |
zigo | fungi: Though it'd be nice to have some consistency checks. For example, if there's python3-all-dev, no need for python-all, and such ... | 15:25 |
samueldmq | fungi: ++ | 15:25 |
*** hurgleburgler1 has quit IRC | 15:25 | |
*** xarses has quit IRC | 15:25 | |
zigo | fungi: It's looking like I'll be voting for absolutely all DefCore projects in this cycle! :P | 15:26 |
*** xarses has joined #openstack-infra | 15:26 | |
fungi | zigo: "voting in all elections" is an exclusive club that most of the infra team also belongs to | 15:26 |
*** e0ne has quit IRC | 15:26 | |
*** ericksonsantos has joined #openstack-infra | 15:26 | |
zigo | fungi: :) | 15:27 |
samueldmq | zigo: cool, let me know when you have fixes for keystone repos | 15:27 |
odyssey4me | fungi hmm, can a var not be set for use for the tox environment? http://logs.openstack.org/70/339770/4/check/gate-project-config-jenkins-project/8e8d772/console.html#_2016-07-12_15_05_00_791085 | 15:27 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Configure a different port for httpd ipxe https://review.openstack.org/340339 | 15:27 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 15:27 |
samueldmq | fungi: clarkb: thanks | 15:27 |
clarkb | dhellmann: I am also going to try filtering out all ht projects that have no code | 15:27 |
clarkb | dhellmann: because apparently that is a non zero set | 15:27 |
*** harlowja_at_home has joined #openstack-infra | 15:27 | |
dhellmann | sigh | 15:27 |
fungi | there are a disappointingly nonzero number of projects that see repo creation in gerrit as a placeholder for work they may never actually get around to starting | 15:29 |
asselin__ | rcarrillocruz, any idea what i'm doing wrong? http://paste.openstack.org/show/530932/ | 15:30 |
*** akshai has quit IRC | 15:31 | |
*** deadnull has quit IRC | 15:31 | |
*** ajmiller has quit IRC | 15:31 | |
rcarrillocruz | i believe you are not passing security_groups var on cli | 15:31 |
rcarrillocruz | thus the parsing fails | 15:32 |
*** armax has joined #openstack-infra | 15:32 | |
*** ajmiller has joined #openstack-infra | 15:32 | |
pabelanger | clarkb: feel free to ping me if you need to offload something in the process | 15:33 |
dhellmann | fungi, clarkb : I suppose asking teams to create something useful in a repo somewhere else would lead to importing lots of untested and/or unreviewed code, so that's not a good solution. How hard is it to remove empty repositories? If we did that annually would that make a difference? | 15:33 |
clarkb | pabelanger: can do, its actually a lot of "make script, run it and keep an eye on it to make sure it doesn't explode" then move on to the next thing | 15:34 |
fungi | dhellmann: it's a small enough percentage to not be worth the effort to try to clean up on their behalf | 15:34 |
clarkb | pabelanger: hwoever I have only been focusing on tox -e py27 -e docs -e releasenotes and -v venv python setup.py build_sphinx so far | 15:35 |
fungi | odyssey4me: i don't know exactly why it's erroring there, though i do see you're unnecessarily re-setting those parameters inside your job templates too | 15:35 |
*** matbu|mtg is now known as matbu | 15:35 | |
clarkb | pabelanger: maybe you want to poke at neutron functional tests and similar tests to that? | 15:35 |
*** arxcruz has quit IRC | 15:35 | |
*** rajinir has joined #openstack-infra | 15:35 | |
*** rdx_matt has joined #openstack-infra | 15:36 | |
*** sdake_ has quit IRC | 15:36 | |
clarkb | pabelanger: I think there is only a small handful of those but parsing zuul layout.yaml should tell you for sure | 15:36 |
*** sdake_ has joined #openstack-infra | 15:36 | |
clarkb | dhellmann: fungi ya its mostly only an annoyance when you try to do these large scale high level "does X work" then you have to filter through and sort out why a great big collection fail | 15:36 |
odyssey4me | fungi so if I understand how it all works correctly, I can remove the 'scenario' and 'toxenv' settings from the job group because the job group inherits them? | 15:36 |
fungi | odyssey4me: oh, that's actually exactly the problem. i'll leave inline comments | 15:36 |
odyssey4me | fungi great, thanks | 15:37 |
clarkb | http://git.openstack.org/cgit/openstack/networking-icc/tree/ is another | 15:37 |
clarkb | actually I think there are a few networking-*s in this bucket | 15:37 |
clarkb | dhellmann: fungi what we can do is put a list together then send them a gentle nudge about using repos if they are still intending to and letting us know if not | 15:37 |
*** dtantsur is now known as dtantsur|afk | 15:37 | |
*** makowals has quit IRC | 15:38 | |
* clarkb uses pep8 fails as a canary | 15:38 | |
*** bauzas is now known as bauzas_notabot | 15:40 | |
asselin__ | rcarrillocruz, I added security groups, but still same error. Seems only name and image are required. | 15:40 |
rcarrillocruz | show me the playbook | 15:40 |
pabelanger | clarkb: sure, I can poke around | 15:40 |
pabelanger | clarkb: can you pastebin your script again? | 15:41 |
*** vincentll has quit IRC | 15:41 | |
jeblair | pabelanger: perhaps we should test the pipelining change on a single zl? | 15:41 |
*** cardeois_ has quit IRC | 15:41 | |
pabelanger | wouldn't mind a review or 2 on 340993 to help close out my reprepro testing | 15:41 |
*** mctaylor has joined #openstack-infra | 15:41 | |
pabelanger | jeblair: I think that is a good plan | 15:42 |
clarkb | pabelanger: http://paste.openstack.org/show/530939/ note the comment at the top with minor setup requirements | 15:42 |
*** cardeois_ has joined #openstack-infra | 15:42 | |
*** mctaylor has quit IRC | 15:42 | |
clarkb | pabelanger: but otherwise running that on one of our images should just work | 15:42 |
anteaya | john-davidge: do you want to try to contact them: https://wiki.openstack.org/wiki/ThirdPartySystems/HP_Networking_CI and point out that they have to follow the requirments: http://docs.openstack.org/infra/system-config/third_party.html#requirements to ensure their gerrit account remains enabled? | 15:42 |
pabelanger | clarkb: will do | 15:42 |
*** jamesmca_ has quit IRC | 15:42 | |
anteaya | john-davidge: if you give that a try and they don't reply or continue to post build started then I can take further action | 15:43 |
jeblair | pabelanger: if you want to graceful stop one of them and manually install that for testing, i'd be ok with that | 15:43 |
clarkb | pabelanger: then I have done for project in cat project run that script | 15:43 |
pabelanger | jeblair: okay, let me set aside some time later today. | 15:43 |
pabelanger | clarkb: ack | 15:43 |
john-davidge | anteaya: Sure, will do :) | 15:43 |
anteaya | john-davidge: thank you, I appreciate your help here :) | 15:44 |
rcarrillocruz | o-k | 15:44 |
rcarrillocruz | finally! | 15:44 |
*** mahatic has quit IRC | 15:44 | |
jeblair | Shrews, notmorgan: i've been deep in crazy for the past couple days; i might have missed requests made of me -- do you need anything? | 15:44 |
rcarrillocruz | Shrews , mordred : i need eyes on https://review.openstack.org/#/c/315713/ | 15:44 |
rcarrillocruz | that should unblock the shade gate | 15:45 |
*** ericksonsantos has left #openstack-infra | 15:46 | |
*** akshai has joined #openstack-infra | 15:46 | |
john-davidge | anteaya: Message sent, will let you know if the big enforcement guns are needed | 15:47 |
*** mikelk has quit IRC | 15:48 | |
*** ericksonsantos has joined #openstack-infra | 15:48 | |
anteaya | john-davidge: very good, hopefully it was a case of education doing what is needed here | 15:48 |
anteaya | thank you | 15:48 |
*** andreas_s has quit IRC | 15:49 | |
*** mestery has quit IRC | 15:49 | |
clarkb | zaro: looks like python-jenkins fails testing on xenial because it can't compile against kerberos | 15:49 |
*** mriedem1 has joined #openstack-infra | 15:49 | |
clarkb | fungi: ^ any idea how that passes on trusty? | 15:49 |
fungi | jeblair: not sure if you saw earlier discussion with odyssey4me and pabelanger, but short version is that it would at least be nice in zuul v3 if workspaces (assuming there still is such a concept) don't embed the job name or repo name in their path. tox uses the full path to bin/python in a virtualenv and if that exceeds BINPRM_BUF_SIZE in /usr/include/linux/binfmts.h (generally 128) then it's | 15:50 |
fungi | truncated | 15:50 |
*** mestery has joined #openstack-infra | 15:50 | |
clarkb | huh krb5-multidev ends up in the trusty jobs somehow | 15:51 |
jeblair | fungi: ah yeah, i bet we can avoid that in v3. | 15:51 |
clarkb | http://logs.openstack.org/08/337208/1/check/gate-python-jenkins-python27/668878c/console.html.gz#_2016-07-04_12_42_32_149226 | 15:51 |
*** mriedem has quit IRC | 15:51 | |
openstackgerrit | Merged openstack-infra/zuul: Support post jobs by supporting rev checkout https://review.openstack.org/293194 | 15:51 |
fungi | clarkb: likely a transitive dep of some package getting installed on trusty but not on xenial? | 15:51 |
clarkb | looks like it may be a dep | 15:51 |
clarkb | fungi: ya | 15:51 |
clarkb | I guess I need to add a patch to include that in fallback requirements | 15:52 |
*** mriedem has joined #openstack-infra | 15:52 | |
fungi | clarkb: yeah, i'm okay extending the fallback list to add analogous xenial support (rather than forcing many projects to add an other-requirements.txt file before we can move to xenial) but for that one repo it might be easier to just switch it to using other-requirements? | 15:53 |
fungi | keep in mind that the fallback list is baked into our images, so will be somewhat delayed taking effect | 15:53 |
*** mahatic has joined #openstack-infra | 15:54 | |
*** mriedem1 has quit IRC | 15:54 | |
*** ddieterly is now known as ddieterly[away] | 15:54 | |
clarkb | thats true, and this run of tests should tell me if there are any other projects that rely on the same dev package | 15:54 |
odyssey4me | fungi with regards to your comment in https://review.openstack.org/#/c/339770/5/jenkins/jobs/projects.yaml - the jobs added are *additional* - in order words the 'functional' tox env must run and the 'functional-ovs' tox env must run | 15:54 |
clarkb | zaro: ^ maybe you can work on other-requirementsing python-jenkins? | 15:55 |
*** gyee has joined #openstack-infra | 15:55 | |
clarkb | fungi: I will likely wait for the entire set to run then compile a "master" list of missing system deps | 15:55 |
openstackgerrit | Doug Hellmann proposed openstack-infra/project-config: add job to apply tags when based on changes in openstack/releases https://review.openstack.org/341058 | 15:55 |
*** psachin has quit IRC | 15:55 | |
odyssey4me | fungi also that job group has other jobs that must also run | 15:55 |
odyssey4me | (docs, linters, announce-release) | 15:56 |
fungi | odyssey4me: oh, yep, i missed the ovs on those. so my last inline comment about redundancy isn't valid | 15:56 |
dhellmann | fungi , clarkb : I would appreciate some help making sure I have ^^ setting up the local workspace properly (and of course anything else needed, but that's the part I know I'm uncomfortable with having done right) | 15:56 |
odyssey4me | fungi ok cool thanks - just making sure I understand things correctly | 15:56 |
zaro | clarkb | 15:56 |
*** arxcruz has joined #openstack-infra | 15:57 | |
zaro | clarkb: will look into that. need to setup xenial first | 15:57 |
*** ddieterly[away] is now known as ddieterly | 15:57 | |
*** ajmiller has quit IRC | 15:57 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 15:57 |
clarkb | zaro: thanks | 15:57 |
*** ifarkas has quit IRC | 15:58 | |
*** ajmiller has joined #openstack-infra | 15:58 | |
rajinir | I need the slave node for debugging failures. Will OFFLINE_NODE_WHEN_COMPLETE='0' work? | 15:59 |
clarkb | rajinir: if you are using nodepool you can `nodepool hold` the instance. If you don't offline the node then more jobs can run on it which is likely not desireable | 15:59 |
jeblair | rajinir: are you using nodepool? | 15:59 |
*** _oanson has quit IRC | 16:00 | |
sshnaidm | clarkb, jeblair hi, can we please merge this today? https://review.openstack.org/#/c/334452/ thanks | 16:00 |
*** pcaruana has quit IRC | 16:00 | |
rajinir | clarkb,thanks will try. Is that using the nodepool cli? | 16:01 |
rajinir | jeblair: yes | 16:01 |
clarkb | rajinir: it is | 16:01 |
rajinir | clarb: awesome , thanks | 16:02 |
rajinir | clarkb: Thank you | 16:02 |
pabelanger | odyssey4me: I've left a comment, and what we need to do to fix the scenario / toxenv settings | 16:02 |
pabelanger | odyssey4me: we need to allow tox -efunc in tox.ini | 16:02 |
Shrews | jeblair: nothing lately from me | 16:03 |
Shrews | rcarrillocruz: looking | 16:03 |
clarkb | sshnaidm: +2 | 16:03 |
odyssey4me | pabelanger that would certainly simplify things - I was trying to avoid the mass tox.ini change | 16:03 |
*** rcernin has quit IRC | 16:03 | |
sshnaidm | clarkb, thanks! | 16:03 |
pabelanger | odyssey4me: Ya, that's the most work, but the right change IMO | 16:03 |
odyssey4me | pabelanger so basically I disconnected the scenario name from the tox env name | 16:03 |
*** jlanoux has quit IRC | 16:04 | |
jeblair | sshnaidm: SpamapS had some good suggestions there -- the first, mirring the names between the method and command line. i noticed you updated your patch to implement that, though i note that while they are closer, they aren't exactly the same. SpamapS, is that what you meant to suggest? | 16:04 |
jeblair | sshnaidm: the other suggestion was adding a basic test -- i'm guessing something that instantiates geard with tcp keepalives just to make sure the code executes (we probably don't need to try to verify that they actually *work*) | 16:05 |
odyssey4me | pabelanger you do realise that it'll be a change to add func as a target to tox, then the job change, then a change to remove functional from tox :/ | 16:05 |
jeblair | sshnaidm: what do you think about adding a test like that? | 16:05 |
sshnaidm | jeblair, mm.. I need to look at current tests and to see if it's possible | 16:06 |
*** daemontool has quit IRC | 16:06 | |
*** dprince has quit IRC | 16:06 | |
*** harlowja_at_home has quit IRC | 16:06 | |
rcarrillocruz | i'm going afk for a bit, enough shade battling | 16:06 |
rcarrillocruz | will be around for infra meeting | 16:06 |
rcarrillocruz | laterz | 16:07 |
*** tonytan_brb has joined #openstack-infra | 16:07 | |
sshnaidm | jeblair, I thought it was related to variable names, not cli args, but no problem to change them too | 16:07 |
pabelanger | odyssey4me: Yup. Going to be alot of churn. I don't think JJB will work on your current patch, lets see what the gate says. otherwise, maybe we just hold off until zuulv3 is done | 16:07 |
jeblair | sshnaidm: i like the cli args as is | 16:07 |
*** ihrachys has quit IRC | 16:07 | |
sshnaidm | jeblair, ok, then won't change :) | 16:07 |
openstackgerrit | Anita Kuno proposed openstack-infra/storyboard: Add example commands for the Tags api https://review.openstack.org/338532 | 16:07 |
jeblair | sshnaidm: my question is, should the class parameters exactly match the cli args, or just be mostly like them (as it is now) | 16:07 |
*** rdx_matt has quit IRC | 16:08 | |
*** tqtran has joined #openstack-infra | 16:08 | |
*** Julien-zte has quit IRC | 16:08 | |
Shrews | rcarrillocruz: +A'd. thx! | 16:09 |
*** sdake_ has quit IRC | 16:09 | |
sshnaidm | jeblair, I think cli args are for people, and class arguments are for library users, so I don't think it *must* be the same | 16:09 |
*** tesseract- has quit IRC | 16:09 | |
*** tonytan4ever has quit IRC | 16:09 | |
*** sdake_ has joined #openstack-infra | 16:10 | |
*** karthikp_ has quit IRC | 16:10 | |
*** simondodsley has joined #openstack-infra | 16:10 | |
jeblair | sshnaidm: *nod*. i'm hoping SpamapS will let us know what his intent was there. | 16:11 |
*** tqtran has quit IRC | 16:12 | |
dmsimard | Oh wow, it never occured to me that http://status.openstack.org/zuul/ and http://zuul.openstack.org/ were different | 16:12 |
* SpamapS stirs | 16:12 | |
dmsimard | http://status.openstack.org/zuul/ is so much more lightweight on my browser | 16:12 |
*** ggillies has quit IRC | 16:13 | |
*** david-lyle has joined #openstack-infra | 16:13 | |
*** hurgleburgler has joined #openstack-infra | 16:14 | |
pabelanger | odyssey4me: left a new suggestion | 16:14 |
*** ggillies has joined #openstack-infra | 16:14 | |
*** javeriak has joined #openstack-infra | 16:14 | |
*** bauzas_notabot is now known as bauzas | 16:15 | |
odyssey4me | pabelanger hmm, that's doable - although the tox builder executes the run-tox script, it doesn't just execute tox -e {env} | 16:16 |
*** jamesmcarthur has joined #openstack-infra | 16:16 | |
*** dizquierdo has quit IRC | 16:18 | |
*** daemontool has joined #openstack-infra | 16:19 | |
*** ajmiller has quit IRC | 16:19 | |
*** ajmiller has joined #openstack-infra | 16:19 | |
SpamapS | sshnaidm: I agree with your assertion that CLI args are different than library args | 16:21 |
*** david-lyle has quit IRC | 16:21 | |
sshnaidm | SpamapS, ok, cool | 16:21 |
*** tidwellr has left #openstack-infra | 16:22 | |
*** armax_ has joined #openstack-infra | 16:22 | |
*** Swami has joined #openstack-infra | 16:22 | |
*** shashank_hegde has joined #openstack-infra | 16:22 | |
SpamapS | sshnaidm: +1'd | 16:23 |
SpamapS | jeblair: ^ | 16:23 |
*** karthikp_ has joined #openstack-infra | 16:23 | |
*** hurgleburgler has quit IRC | 16:23 | |
*** sandanar_ has quit IRC | 16:23 | |
*** notnownikki has joined #openstack-infra | 16:23 | |
jeblair | SpamapS: okay, so you were going for "approximate" rather than "be identical"? | 16:23 |
*** gongysh has quit IRC | 16:23 | |
pabelanger | odyssey4me: right, we'd call /usr/local/jenkins/slave_scripts/run-tox.sh {scenario} directly now | 16:24 |
pabelanger | or | 16:24 |
*** sdake_ has quit IRC | 16:24 | |
*** armax has quit IRC | 16:24 | |
*** armax_ is now known as armax | 16:24 | |
pabelanger | /usr/local/jenkins/slave_scripts/run-tox.sh functional | 16:24 |
*** sandanar has quit IRC | 16:24 | |
*** d0ugal has quit IRC | 16:24 | |
*** jpich_ has quit IRC | 16:24 | |
SpamapS | jeblair: correct. The way they were written before, they seemed oddly skewed. But now they're at least the same letters (or at least, my brain thinks they are) | 16:25 |
odyssey4me | pabelanger ok, adjustment incoming | 16:26 |
*** jtomasek__ has quit IRC | 16:26 | |
*** jlk` is now known as jlk | 16:26 | |
*** jlk has quit IRC | 16:27 | |
*** jlk has joined #openstack-infra | 16:27 | |
openstackgerrit | Matt Riedemann proposed openstack-infra/project-config: Make gate-tempest-dsvm-postgres-full voting again https://review.openstack.org/341073 | 16:27 |
mriedem | anteaya: clarkb: ^ | 16:27 |
*** hurgleburgler has joined #openstack-infra | 16:27 | |
sshnaidm | SpamapS, jeblair so we have an agreement? | 16:29 |
pabelanger | rcarrillocruz: have a minute for a review? 340993 | 16:29 |
*** makowals has joined #openstack-infra | 16:30 | |
jeblair | sshnaidm: yep. do you want to add a test? | 16:31 |
vponomaryov | fungi, tristanC: Hello, I need help in procedure of security bug fix proposal. "manila-ui" project does not have vulnerability managed governance tag, but has currently "private" security issue reported with existing fix for it. So, what our, "manila-ui" maintainers, proper actions to do? | 16:31 |
sshnaidm | jeblair, as I looked now on tests, this server is not tested anywhere, it's itself: "A simple gearman server implementation for testing (not for production use)" | 16:32 |
clarkb | fungi: ttx tristanC ^ I think vponomaryov's question is somethign you may know the answer to | 16:32 |
sshnaidm | jeblair, so testing it will require a pretty big overhead in test code, but no real value, because it's a test itself.. | 16:32 |
sshnaidm | we're bad risky guys that use it :) | 16:33 |
*** trown is now known as trown|lunch | 16:33 | |
jeblair | sshnaidm: works for me :) | 16:33 |
*** savihou has quit IRC | 16:34 | |
*** Rockyg has joined #openstack-infra | 16:34 | |
fungi | vponomaryov: the official process we have documented at https://security.openstack.org/vmt-process.html also serves as a guideline for handling of vulnerabilities outside of openstack vmt oversight. there are a few caveats though | 16:34 |
sshnaidm | jeblair, SpamapS, clarkb thank for your input guys! | 16:34 |
anteaya | mriedem: +2, thank you | 16:35 |
jeblair | sshnaidm: thanks for the patch :) we should probably start using it.... | 16:35 |
fungi | vponomaryov: also, we're (vmt members) happy to consult on vulnerability reports throughout the community and guide you through some of the more touchy parts of that process | 16:35 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 16:36 |
*** david-lyle has joined #openstack-infra | 16:36 | |
*** e0ne has joined #openstack-infra | 16:36 | |
sshnaidm | jeblair, yeah, if zuul erases vms not gracefully, you have a risk to have a lot of stuck connection and breaking fd limits | 16:36 |
fungi | vponomaryov: the main caveats are that the published vulnerability won't have an "ossa" number assigned, and you'll need vmt assistance to send advance notification to downstream stakeholders if you need to do so under an embargo (depending on your comfort level and the severity of the vulnerability) | 16:37 |
*** asettle has quit IRC | 16:37 | |
fungi | vponomaryov: otherwise pretty much all of that process should be reusable by your project's bug team | 16:38 |
*** rkukura has quit IRC | 16:38 | |
*** mrmartin has quit IRC | 16:38 | |
openstackgerrit | Merged openstack-infra/gear: Support TCP keepalives in geard https://review.openstack.org/334452 | 16:38 |
odyssey4me | pabelanger you like ? https://review.openstack.org/339770 :) | 16:39 |
pabelanger | odyssey4me: looking | 16:39 |
*** tqtran has joined #openstack-infra | 16:39 | |
fungi | vponomaryov: oh, also you won't propose a yaml version of the advisory to the openstack/ossa repo, but instead will need to plug your information into a copy of the restructured text we use to send our advisories to various mailing lists | 16:39 |
*** sdake_ has joined #openstack-infra | 16:40 | |
*** shashank_hegde has quit IRC | 16:40 | |
vponomaryov | fungi: I am not sure whether OSSA is needed, but "vmt assistance to send advance notification to downstream stakeholders" is right thing to do, IMHO | 16:40 |
*** asettle has joined #openstack-infra | 16:41 | |
fungi | vponomaryov: you can see an example by looking at a recent official advisory like https://security.openstack.org/vmt-process.html and clicking on the "show source" link in the left hand column | 16:41 |
*** AJaeger has joined #openstack-infra | 16:41 | |
fungi | that's basically what we send in the public announcements | 16:41 |
anteaya | anyone else willing to review mriedem's patch to make the progress-full job voting again? https://review.openstack.org/#/c/341073/1 | 16:41 |
openstackgerrit | Jon-Paul Sullivan (j3p0uk) proposed openstack-infra/zuul: Make the cloner strategy clearer https://review.openstack.org/177201 | 16:42 |
pabelanger | odyssey4me: minor nit | 16:42 |
fungi | vponomaryov: er, i meant a recent advisory like https://security.openstack.org/ossa/OSSA-2016-008.html | 16:42 |
*** eharney has joined #openstack-infra | 16:42 | |
*** thorst is now known as thorst_afk | 16:42 | |
*** lucasagomes is now known as lucas|afk | 16:42 | |
sshnaidm | jeblair, do you know maybe - how is it going the versioning/releasing geard? when can I get the new version from PyPi? | 16:43 |
AJaeger | jeblair: thanks for merging Sachi's change for zuul (https://review.openstack.org/#/c/293194/)- what does this mean for contraints? Can we use them now in post jobs - and how exactly? | 16:43 |
pabelanger | AJaeger: did you submit your lightning talk for bindep? | 16:43 |
fungi | vponomaryov: anyway, feel free to contact those of us listed at https://security.openstack.org/index.html#how-to-report-security-issues-to-openstack or subscribe https://launchpad.net/~openstack-vuln-mgmt to the bug | 16:43 |
AJaeger | pabelanger: yes, I did - since nobody answered my email, I sent something yesterday | 16:43 |
AJaeger | pabelanger: thanks for asking! | 16:44 |
pabelanger | AJaeger: great, I just did one for nodepool and dibs | 16:44 |
jeblair | sshnaidm: it might be nice to land https://review.openstack.org/184265 before releasing it | 16:44 |
jeblair | clarkb: ^ if you feel like another gear review | 16:44 |
jeblair | (also parent patch) | 16:44 |
AJaeger | pabelanger: great! | 16:45 |
openstackgerrit | Khai Do proposed openstack-infra/system-config: Configure its-storyboard plugin https://review.openstack.org/330925 | 16:45 |
openstackgerrit | Merged openstack-infra/shade: Add floating IPs to server dict ourselves https://review.openstack.org/315713 | 16:46 |
jeblair | AJaeger: i don't know exactly what that will look like | 16:46 |
*** asettle has quit IRC | 16:46 | |
*** karthikp_ has quit IRC | 16:46 | |
*** david-lyle_ has joined #openstack-infra | 16:46 | |
*** garyk has quit IRC | 16:46 | |
greghaynes | pabelanger: nice! | 16:46 |
clarkb | jeblair: yes I need a break from trolling failed test logs | 16:46 |
*** asettle has joined #openstack-infra | 16:46 | |
*** karthikp_ has joined #openstack-infra | 16:47 | |
dmsimard | Is there a way for a job to tell if it's running a parent or a child job in Zuul ? i.e with the config in http://paste.openstack.org/show/530955/ .. I'd basically like to re-use the child job in different contexts and just add an "<if running as child>do this</if running as child>". | 16:47 |
jeblair | AJaeger: but once new images are built with the new zuul-cloner, it can be used in post jobs. we install from master, so no zuul release needed. | 16:47 |
AJaeger | samueldmq, clarkb : you had a question on other-requirements? Ideally, the normal environment is run-time environment, and the test adds the packages needed for testing. | 16:47 |
pabelanger | AJaeger: if you mind adding 340993 to your review list | 16:47 |
AJaeger | pabelanger: will do | 16:47 |
AJaeger | dmsimard, clarkb: Is there a review you want comments on? | 16:48 |
* AJaeger glances over eavesdrop backscroll | 16:48 | |
AJaeger | pabelanger: will do | 16:48 |
dmsimard | AJaeger: this is in another zuul instance, not openstack-infra's | 16:48 |
samueldmq | AJaeger: so runtime env is always a superset of dev env | 16:48 |
*** david-lyle has quit IRC | 16:48 | |
jeblair | dmsimard: not directly, however, most child jobs are children because of some dependency on the product of a parent job, so rather than assuming that dependency, you could check for it. | 16:49 |
AJaeger | samueldmq: as clarkb said it. I see runtime as subset of test environment - for testing, you need additional packages installed. | 16:49 |
samueldmq | AJaeger: and we differentiate the test specific reqs with [test] section in other-requirements, right? | 16:49 |
dmsimard | jeblair: the way we thought of doing it was to query the Zuul status.json and find if there is a relationship for the same patch | 16:49 |
openstackgerrit | Jeremy Stanley proposed openstack-infra/infra-specs: Add spec for unified message bus https://review.openstack.org/329115 | 16:49 |
dmsimard | was hoping there'd be a better way | 16:49 |
dmsimard | but yes, in this particular case, we're interested in an artifact of the parent job | 16:50 |
jeblair | dmsimard: all jobs in a build set share a zuul ref | 16:50 |
samueldmq | AJaeger: that's correct, I've put it in the opposite order in my explanation, gotcha | 16:50 |
AJaeger | jeblair: ok, do we need to change existing post jobs to make use of the change zuul-cloner? | 16:50 |
clarkb | jeblair: done, that was a simple change | 16:50 |
AJaeger | samueldmq: yes, [test] - as documented. Have a look at storyboard repo for an example | 16:50 |
jeblair | dmsimard: the zuul ref can be used to uniquely identify artifacts from the same build set | 16:51 |
jeblair | dmsimard: please don't fetch status.json :) | 16:51 |
openstackgerrit | Jeremy Stanley proposed openstack-infra/infra-specs: Firehose, a unified message bus for Infra services https://review.openstack.org/329115 | 16:52 |
samueldmq | AJaeger: https://github.com/openstack-infra/storyboard/blob/master/other-requirements.txt ? | 16:52 |
AJaeger | zigo, dmsimard : Initially using [test] was broken, so first repos did not use [test] at all. Feel free to update those repos to make proper use of [test] | 16:52 |
*** Daisy_ has joined #openstack-infra | 16:52 | |
samueldmq | AJaeger: I can't see the use of [test] there | 16:52 |
dmsimard | jeblair: hm, the problem I'm facing is that I have to find the zuul_uuid of the parent job -- because that uuid is used in the artifact location and the child job does not know it | 16:52 |
clarkb | dhellmann: is the CalledProcessError for releasenotes the one you are aware of? | 16:52 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement optional scenario setting for Ansible jobs https://review.openstack.org/339770 | 16:52 |
dmsimard | jeblair: since the zuul_uuid is unique per job | 16:52 |
AJaeger | dmsimard: my change hasn't merged - use https://review.openstack.org/#/c/335826/1/other-requirements.txt | 16:52 |
*** notnownikki has quit IRC | 16:53 | |
AJaeger | dmsimard: I forgot that it's still under review ;( | 16:53 |
dmsimard | AJaeger: have you been meaning to ping someone else ? I'm not sure what you're on about | 16:53 |
samueldmq | AJaeger: dmsimard: cool, I looked at that patch ^ thanks | 16:53 |
jeblair | dmsimard: i'd suggest using the zuul ref rather than the uuid then | 16:54 |
*** javeriak_ has joined #openstack-infra | 16:54 | |
*** javeriak has quit IRC | 16:54 | |
AJaeger | dmsimard: I meant samueldmq - not sure how I screwed up. Sory. | 16:54 |
odyssey4me | pabelanger done | 16:54 |
dmsimard | jeblair: so upload artifacts at a location based on the ref rather than the uuid ? | 16:54 |
samueldmq | AJaeger: np at all | 16:54 |
samueldmq | AJaeger: thanks for the additional clarifications | 16:55 |
*** Apoorva has joined #openstack-infra | 16:55 | |
dmsimard | jeblair: that ref is unique within a single job set, right ? if I do, for example, a recheck, a new ref will be generated ? | 16:55 |
*** Apoorva has quit IRC | 16:55 | |
*** zhurong has joined #openstack-infra | 16:56 | |
*** Daisy_ has quit IRC | 16:56 | |
*** derekh has quit IRC | 16:56 | |
*** dprince has joined #openstack-infra | 16:57 | |
*** rkukura has joined #openstack-infra | 16:57 | |
openstackgerrit | Kaiyan Sheng proposed openstack-infra/project-config: [Monasca-agent] Remove check-requirements check job for monasca-agent https://review.openstack.org/341091 | 16:58 |
jeblair | dmsimard: the ref is unique for a build set, which is a set of running jobs for a change in a given configuration in a pipeline. all the jobs running for a change at the same time share the buildset; you get a new buildset if zuul reorders the queue, and you get a new buildset each time a change is enqueued. | 16:59 |
jeblair | dmsimard: (long winded way of saying 'yes' to your question, i think :) | 16:59 |
*** ajmiller has quit IRC | 16:59 | |
dmsimard | jeblair: ack, thanks | 16:59 |
*** ajmiller has joined #openstack-infra | 17:00 | |
*** Apoorva has joined #openstack-infra | 17:00 | |
openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Bindep: Use python3-all https://review.openstack.org/341093 | 17:00 |
*** xenogear has joined #openstack-infra | 17:00 | |
AJaeger | zigo, fungi, what do you think of this cleanup for python3-all ? ^ | 17:00 |
*** sarob has joined #openstack-infra | 17:01 | |
*** jamesmcarthur has quit IRC | 17:02 | |
*** zhurong has quit IRC | 17:02 | |
*** zhurong has joined #openstack-infra | 17:03 | |
zaro | clarkb, fungi : I've got gerrit delete-project plugin ready to go if you need to cleanup repos, https://review.openstack.org/#/c/319400/ | 17:03 |
AJaeger | infra-root, has anybody helped sdake to upload current images to rax-ord? See https://review.openstack.org/339776 - it showed that we still have old rax-ord images without the updated repository location there. Help would be welcome | 17:03 |
*** mrmartin has joined #openstack-infra | 17:04 | |
*** armax_ has joined #openstack-infra | 17:04 | |
pabelanger | AJaeger: ya, I missed that one last night | 17:04 |
pabelanger | I'll be confirming again tonight | 17:04 |
*** matthewbodkin has quit IRC | 17:04 | |
AJaeger | pabelanger: thanks! | 17:05 |
clarkb | zaro: I think first step is to just let them know that the resources are still available to them and give them a chacne to say they don't want them | 17:05 |
zaro | clarkb: yeah ++ | 17:05 |
dmsimard | jeblair: hm, I guess we're bound to the uuid location as we're currently using swift uploader and there's https://github.com/openstack-infra/project-config/blob/master/jenkins/scripts/zuul_swift_upload.py#L467 | 17:05 |
pabelanger | AJaeger: actually, it is updated already | 17:05 |
*** rdx_matt has joined #openstack-infra | 17:06 | |
clarkb | nodepool tsets seem to be completely broken on xenial | 17:06 |
dmsimard | jeblair: but we can override that location | 17:06 |
clarkb | due to mysql things | 17:06 |
pabelanger | AJaeger: so each region should be good for ubuntu-trusty now | 17:06 |
AJaeger | sdake: let's recheck again ^ | 17:06 |
*** akshai has quit IRC | 17:06 | |
AJaeger | thanks, pabelanger | 17:06 |
pabelanger | AJaeger: which review are you testing? | 17:06 |
zaro | fungi: RE:https://review.openstack.org/#/c/334714/ gerrit module is not part of puppet-openstackci so are third parties actually using it? | 17:07 |
*** akshai has joined #openstack-infra | 17:07 | |
*** armax has quit IRC | 17:07 | |
*** armax_ is now known as armax | 17:07 | |
AJaeger | pabelanger: 339776 | 17:08 |
zaro | if so can we identify and add appropriate reviewers to get feedback. otherwise i can just make it configurable if it makes this go in faster | 17:08 |
pabelanger | dmsimard: you should be able to update your copy of http://git.openstack.org/cgit/openstack-infra/project-config/tree/zuul/openstack_functions.py#n18 for the LOG_PATH you need | 17:08 |
*** karthikp_ has quit IRC | 17:09 | |
dmsimard | pabelanger: oh, that's much cleaner than what I was going to do, thanks for that | 17:09 |
fungi | zaro: i don't think it has as many downstream consumers as modules openstackci depends on, but it definitely has some | 17:09 |
zaro | fungi: would making it configurable satisfy your concern? | 17:09 |
anteaya | zaro: I have no idea how many downstream folks are consuming the gerrit module, the only way I would know is if you post to -infra, asking for them to show themselves and review the patch if they care | 17:09 |
fungi | zaro: i understood hpe/gozer was using it, for example | 17:09 |
*** sarob has quit IRC | 17:09 | |
*** karthikp_ has joined #openstack-infra | 17:09 | |
fungi | zaro: and yes, making it configurable would solve that | 17:09 |
zaro | ok, then i'll just puppet it up configurable. | 17:10 |
fungi | zaro: that way we can enable git gc from our system-config repo but others who are using that module continue to use git repack instead | 17:10 |
dmsimard | pabelanger: when does that openstack_functions script run ? | 17:11 |
pabelanger | dmsimard: from zuul/layout.yaml | 17:11 |
*** ihrachys has joined #openstack-infra | 17:11 | |
pabelanger | dmsimard: it gets loaded when zuul starts IIRC | 17:12 |
dmsimard | pabelanger: I mean, I essentially want to remap LOG_PATH to "${BASE_LOG_PATH}/${ZUUL_PIPELINE}/${JOB_NAME}/${ZUUL_REF}" | 17:12 |
dmsimard | do I access to those variables ? | 17:12 |
dmsimard | with os.getenv or whatever | 17:12 |
pabelanger | and then we use some regex to call the set_node_options function | 17:12 |
pabelanger | Umm | 17:12 |
dmsimard | Right now LOG_PATH is ${BASE_LOG_PATH}/${ZUUL_PIPELINE}/${JOB_NAME}/${ZUUL_UUID} | 17:13 |
pabelanger | dmsimard: http://git.openstack.org/cgit/openstack-infra/project-config/tree/zuul/layout.yaml#n1057 is how it is setup | 17:13 |
*** PollyZ has quit IRC | 17:13 | |
pabelanger | do, I am not sure which jobs you want to override it for | 17:13 |
pabelanger | there is a few ways to handle it | 17:13 |
pabelanger | you can do it in JJB | 17:13 |
pabelanger | or from zuul directly | 17:13 |
dmsimard | pabelanger: yeah I was going to jenkins way | 17:14 |
dmsimard | or, err, well, JJB | 17:14 |
openstackgerrit | Merged openstack-infra/project-config: Enable reprepro-release-deb-openstack https://review.openstack.org/340993 | 17:14 |
*** pfallenop has quit IRC | 17:15 | |
*** jtomasek__ has joined #openstack-infra | 17:17 | |
*** pfallenop has joined #openstack-infra | 17:17 | |
*** degorenko is now known as _degorenko|afk | 17:17 | |
*** sdake_ has quit IRC | 17:17 | |
dkehn | fungi, is there a openstack-infra repo that is used to build the the final release packages for things like neutron, etc.? | 17:17 |
*** gfreeman has joined #openstack-infra | 17:18 | |
anteaya | dkehn: there are release manager tools, could that be what you are looking for? | 17:18 |
anteaya | dkehn: http://git.openstack.org/cgit/openstack-infra/release-tools/ | 17:19 |
jesusaur | clarkb: how busy is your week looking? could we possibly find some time to switch to the log_processor project (I'm realizing that I dropped the ball on that a while ago) | 17:20 |
dkehn | anteaya, so when the project for release are build is there a set of tools , in the openstack-info repo tree, for example that are used to package up the rpms, etc for real release? | 17:20 |
dkehn | anteaya, ok, let me looky | 17:21 |
anteaya | I don't think we package rpms | 17:21 |
jesusaur | but the good news is that https://review.openstack.org/254312 is the only outstanding commit | 17:21 |
dkehn | anteaya, thxs tons | 17:21 |
clarkb | jesusaur: this week is a milestone and I am trying to finish up prep for the xenial switch so we can switchover soon after milestone | 17:21 |
anteaya | dkehn: I think the artifact form we create are tarballs, then folks that want other package forms use the tarballs | 17:21 |
clarkb | jesusaur: I may have time late in the week depending onhow prep goes | 17:21 |
anteaya | dkehn: and you are welcome tons | 17:22 |
*** zhurong has quit IRC | 17:22 | |
clarkb | jesusaur: so far it is going pretty well. Nothing too crazy has been run into so I may have time | 17:22 |
*** jtomasek__ has quit IRC | 17:22 | |
AJaeger | dkehn, there's a separate project for RPMs with their own channel... | 17:22 |
AJaeger | dkehn: the release team only takes care of python packages and that's basically invoking the usual sdist commands and uploading | 17:22 |
fungi | dkehn: yeah, what anteaya says. we publish python sdist tarballs and wheels from every tag. various linux distro package maintainers have their own tools and workflows for making the packages of those for their distros--we don't have a hand in it currently | 17:24 |
dkehn | AJaeger, thx | 17:24 |
jesusaur | clarkb: alright, ping me if you want to make the switch | 17:25 |
*** locust has joined #openstack-infra | 17:25 | |
locust | hello, can anybody please give me link on how to deploy my local "Cloud" that supposed to be opperated by nodepool? | 17:26 |
*** dizquierdo has joined #openstack-infra | 17:26 | |
locust | and it is all in a context of OpenStack CI | 17:26 |
locust | thank you | 17:26 |
*** samueldmq has quit IRC | 17:27 | |
*** garyk has joined #openstack-infra | 17:28 | |
*** mrmartin has quit IRC | 17:28 | |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 17:28 |
fungi | locust: we (for the upstream openstack ci) don't actually run a cloud nodepool interacts with (yet). we just use accounts at various public cloud providers. so it's at least not necessary in all cases that you run a copy of openstack to use nodepool | 17:29 |
*** dizquierdo is now known as dizquierdo_afk | 17:30 | |
*** deadnull has joined #openstack-infra | 17:31 | |
*** hurgleburgler has quit IRC | 17:31 | |
*** hurgleburgler has joined #openstack-infra | 17:31 | |
fungi | locust: if you want to run your own openstack cloud and use that with nodepool, you would i suppose deploy openstack via whatever means you're most comfortable with (there are projects which deploy it with ansible, puppet, chef, salt, in containers, bootstrapping it with itself and orchestrating with heat, pick your poison) | 17:31 |
*** karthikp_ has quit IRC | 17:31 | |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 17:32 |
jeblair | dmsimard: i thought we were not using the swift uploader for anything right now | 17:32 |
anteaya | thank you AJaeger | 17:33 |
*** ddieterly is now known as ddieterly[away] | 17:33 | |
*** ddieterly[away] is now known as ddieterly | 17:34 | |
jeblair | dmsimard: what zuul are you talking about? openstack's zuul or a third-party? | 17:34 |
openstackgerrit | Chris Denneen proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options https://review.openstack.org/312121 | 17:34 |
jeblair | dmsimard: because pabelanger said "your copy of openstack_functions" which only makes sense as a response for a question about a third-party ci | 17:34 |
*** makowals has quit IRC | 17:34 | |
*** makowals has joined #openstack-infra | 17:34 | |
*** makowals has quit IRC | 17:35 | |
*** PollyZ has joined #openstack-infra | 17:35 | |
pabelanger | jeblair: since dmsimard was asking, it is likely Software Factory and review.rdoproject.org | 17:35 |
*** makowals has joined #openstack-infra | 17:36 | |
openstackgerrit | Chris Denneen proposed openstack-infra/jenkins-job-builder: I've rebased my change... see if that helps https://review.openstack.org/341110 | 17:36 |
*** garyk has quit IRC | 17:37 | |
*** garyk has joined #openstack-infra | 17:37 | |
locust | fungi: you mean I can use devstack or rdo too? | 17:37 |
dhellmann | clarkb : hmm, no, I think we were seeing encoding errors. have a link? | 17:37 |
clarkb | dhellmann: its from my local tests on xenial will give you a copy of an example | 17:38 |
dhellmann | k | 17:38 |
fungi | locust: well, i'd advise against devstack since it's not designed to run long-term or preserve configuration changes | 17:38 |
*** shashank_hegde has joined #openstack-infra | 17:38 | |
clarkb | dhellmann: http://paste.openstack.org/show/531005/ | 17:38 |
fungi | locust: though you could certainly test nodepool out against a devstack-deployed openstack (and in fact, that is precisely what our integration testing for nodepool does) | 17:39 |
dhellmann | clarkb : is that a result of running the releasenotes job under python 3.5? | 17:39 |
dmsimard | jeblair: this is in the review.rdoproject.org environment, yes | 17:39 |
fungi | locust: but yes, you could use nodepool with rdo or mos or whatever distro of openstack you like too | 17:40 |
*** rdx_matt has quit IRC | 17:40 | |
locust | fungi: yep, you are 100% right | 17:40 |
*** PollyZ has quit IRC | 17:41 | |
jeblair | dmsimard, pabelanger: gotcha, thanks | 17:41 |
*** dprince has quit IRC | 17:41 | |
openstackgerrit | Eric Brown proposed openstack-infra/project-config: Enable py35 voting on vmware-nsx https://review.openstack.org/340708 | 17:41 |
*** asettle has quit IRC | 17:41 | |
clarkb | dhellmann: no the traceback has python2.7 in it | 17:41 |
locust | fungi: after I deploy for instance RDO openstack, I have to pass credentianl to it to ansible along side with nodepool | 17:41 |
openstackgerrit | Merged openstack-infra/project-config: Make gate-tempest-dsvm-postgres-full voting again https://review.openstack.org/341073 | 17:42 |
dhellmann | oh, weird | 17:42 |
*** akshai has quit IRC | 17:42 | |
*** trown|lunch is now known as trown | 17:42 | |
clarkb | dhellmann: tox should default to python2.7 unless you tell it to us a different version | 17:42 |
dhellmann | well, it defaults to the version of python where it's installed, but yeah | 17:43 |
*** dprince has joined #openstack-infra | 17:43 | |
fungi | locust: you would configure clouds.yaml for nodepool to have the credentials for a tenant/project you created for it in your openstack deployment, yes | 17:43 |
dhellmann | clarkb: ok, this is a new error I haven't seen before. I'll have to see about setting up a test system. this is 16.04? | 17:44 |
fungi | locust: in fact, you could run your nodepool server on a virtual machine in one tenant in your cloud, and have it manage nodes in a separate tenant if you wanted, as long as your api endpoints are reachable from the instance you're using for nodepool | 17:44 |
clarkb | dhellmann: yup 16.04 xenial | 17:44 |
dhellmann | ok, I'll put that on my todo list to test out | 17:45 |
*** akshai has joined #openstack-infra | 17:45 | |
clarkb | dhellmann: 2.3.1 imported from /usr/local/lib/python2.7/dist-packages/tox/__init__.pyc is the tox --version output on my test box | 17:45 |
AJaeger | clarkb: can you run the git command directly? "git log --simplify-by-decoration --pretty="%d" origin/stable/liberty" ? | 17:45 |
clarkb | which is how we install it on the infra images ( I am using an infra image) | 17:45 |
clarkb | AJaeger: ya I can swing around once I am done classifying the current set of errors | 17:45 |
*** jlk has left #openstack-infra | 17:45 | |
*** thorst_afk is now known as thorst | 17:45 | |
dhellmann | clarkb : ack. it might be tomorrow before I can get something similar set up, but I'll look at this one if you want to start looking at the others | 17:46 |
*** mrmartin has joined #openstack-infra | 17:46 | |
*** psilvad has quit IRC | 17:46 | |
*** puiterwijk is now known as puiterwijk|otg | 17:47 | |
*** puiterwijk|otg is now known as puiterwijk|outag | 17:47 | |
*** puiterwijk|outag is now known as puiterwijk | 17:49 | |
*** asettle has joined #openstack-infra | 17:49 | |
*** asettle has quit IRC | 17:49 | |
*** PollyZ has joined #openstack-infra | 17:49 | |
clarkb | dhellmann: sounds good thanks | 17:49 |
*** asettle has joined #openstack-infra | 17:49 | |
clarkb | dhellmann: I will also provide a full list of errors and the sanitized list with things scrubbed out that I believe are non issues | 17:50 |
*** john-davidge has quit IRC | 17:50 | |
*** garyk has quit IRC | 17:51 | |
*** andymaier has joined #openstack-infra | 17:51 | |
*** psilvad has joined #openstack-infra | 17:52 | |
*** tidwellr has joined #openstack-infra | 17:52 | |
tidwellr | Hi, is there someone here who can take a look at (and hopefully approve) https://review.openstack.org/#/c/340222 ? | 17:53 |
*** pvaneck has joined #openstack-infra | 17:53 | |
*** psilvad has quit IRC | 17:54 | |
*** psilvad has joined #openstack-infra | 17:54 | |
*** PollyZ has quit IRC | 17:54 | |
*** tonytan_brb has quit IRC | 17:54 | |
*** ddieterly is now known as ddieterly[away] | 17:55 | |
*** psilvad has quit IRC | 17:55 | |
AJaeger | tidwellr: it needs neutron infra liason +1 first. Ask armax, dougwig or mestery for a +1 - and then somebody will approve. But not before. | 17:56 |
anteaya | tidwellr: can you get armax or dougwig to at least say they know the patch exists? | 17:56 |
anteaya | oh yeah or mestery | 17:56 |
clarkb | dtroyer: openstackclient fails tox -epy27 on xenial with AttributeError: 'module' object has no attribute 'select_auth_plugin' | 17:56 |
anteaya | thanks AJaeger | 17:56 |
dougwig | AJaeger: looking | 17:56 |
* armax looks | 17:56 | |
*** tosky has quit IRC | 17:56 | |
anteaya | the whole gang is here | 17:56 |
AJaeger | thanks, anteaya ;) | 17:56 |
anteaya | :) | 17:56 |
* AJaeger shouldn't have awakened all of them ;) | 17:56 | |
armax | AJaeger: what did you do? | 17:56 |
armax | AJaeger: :) | 17:56 |
*** electrofelix has quit IRC | 17:57 | |
anteaya | it is nice to see them :) | 17:57 |
dougwig | AJaeger: the project with the change in question needs a PROJECTS entry??? | 17:57 |
AJaeger | armax: just helping tidwellr :) | 17:57 |
*** psilvad has joined #openstack-infra | 17:57 | |
*** fawadkhaliq has joined #openstack-infra | 17:57 | |
dougwig | i thought that was implicit. | 17:57 |
armax | AJaeger: too much help at once may have the opposite effect! | 17:57 |
AJaeger | armax: I have no interest in the change - besides that it's in our queue ;) | 17:58 |
*** e0ne has quit IRC | 17:58 | |
*** andymaier has quit IRC | 17:58 | |
armax | AJaeger: it’s all god | 17:58 |
armax | good | 17:58 |
armax | AJaeger: thanks | 17:58 |
*** andymaier has joined #openstack-infra | 17:58 | |
AJaeger | ;) | 17:58 |
*** Apoorva_ has joined #openstack-infra | 17:59 | |
tidwellr | AJaegar: I'm fumbling around just trying to get some jobs to run in the gate for that project, maybe a PROJECTS entry isn't necessary, I have no idea what I'm doing but it seems like getting this merged may help my cause? | 17:59 |
openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Use strategy free for git servers https://review.openstack.org/341121 | 17:59 |
*** ajmiller has quit IRC | 18:00 | |
pabelanger | some ansible performance tuning on puppetmaster.o.o, would be interested in some feedback^. I've limited it to our git servers for the moment | 18:00 |
*** ajmiller has joined #openstack-infra | 18:00 | |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 18:01 |
AJaeger | tidwellr: it's at least consistent with other projects ;) | 18:01 |
clarkb | dhellmann: just thinking about that fail is it possible I need to do more setup than git clone and git checkout master and tox -ereleasenotes? | 18:03 |
*** Apoorva has quit IRC | 18:03 | |
clarkb | dhellmann: since it seems to be complaining about older branches, clone won't setup tracking branches by default. However it is refering to origin | 18:03 |
*** rbrndt has quit IRC | 18:03 | |
clarkb | I wonder if using the on host git cache is breaking this somehow | 18:03 |
AJaeger | clarkb: it should be fine since it refers to origin | 18:04 |
pabelanger | zigo: mordred: okay, here is our first iteration on reprepro: http://mirror.dfw.rax.openstack.org/debian-openstack/ | 18:04 |
clarkb | AJaeger: dhellmann I actually think that could be the issue, I will test that in a bit | 18:04 |
pabelanger | zigo: mordred: I believe we need to have a discussion on the components section. Right now, it contains jessie main (mitaka) | 18:05 |
pabelanger | zigo: mordred: that should likely be jessie/mitaka | 18:05 |
*** ramishra_ has quit IRC | 18:05 | |
pabelanger | but will wait for your feedback | 18:05 |
*** hurgleburgler has quit IRC | 18:06 | |
clarkb | ya my local clone doesn't have those branches | 18:06 |
*** PollyZ has joined #openstack-infra | 18:07 | |
AJaeger | clarkb: it worked for me locally but perhaps my clone is done differently | 18:07 |
*** ramishra has joined #openstack-infra | 18:07 | |
*** amotoki has quit IRC | 18:07 | |
armax | AJaeger: if I can trouble with https://review.openstack.org/#/c/340612/, that should fix the failure on one of the jobs for the neutron periodic queue | 18:08 |
*** eharney has quit IRC | 18:08 | |
clarkb | AJaeger: its because I am cloning from /opt/git/ on our images | 18:08 |
armax | since you summoned me, in the channel :) | 18:08 |
clarkb | AJaeger: I need to clone from there, then change the remote then update | 18:08 |
armax | mordred, notmorgan: anything you ended up discovering about the shade/neutron issue? | 18:08 |
AJaeger | armax: will check;) | 18:08 |
armax | AJaeger: in your own time | 18:09 |
armax | it’s already late your end | 18:09 |
AJaeger | clarkb: So, you didn't do all steps we do in the gate? Ok, that explains it... | 18:09 |
*** dizquierdo_afk is now known as dizquierdo | 18:09 | |
AJaeger | armax: why is that change not needed for oslo? Are they not using constraints? | 18:10 |
armax | AJaeger: not in that case since it’s testing the master version of the library | 18:11 |
*** sdake_ has joined #openstack-infra | 18:11 | |
armax | AJaeger: that’s the very purpose of the job | 18:11 |
*** akshai has quit IRC | 18:11 | |
*** jamesmcarthur has joined #openstack-infra | 18:11 | |
AJaeger | ah... | 18:11 |
AJaeger | armax: +2 - now find somebody else ;) | 18:12 |
openstackgerrit | James E. Blair proposed openstack-infra/infra-specs: Update Xenial work items formatting https://review.openstack.org/341126 | 18:12 |
jeblair | clarkb: ^ | 18:12 |
armax | AJaeger: thank you sire | 18:12 |
* AJaeger calls it a day and waves good-bye | 18:12 | |
*** andymaier has quit IRC | 18:12 | |
*** hurgleburgler has joined #openstack-infra | 18:12 | |
anteaya | bye AJaeger | 18:12 |
*** ajmiller has quit IRC | 18:13 | |
anteaya | thank you | 18:13 |
*** ajmiller has joined #openstack-infra | 18:13 | |
*** akshai has joined #openstack-infra | 18:13 | |
*** gouthamr has quit IRC | 18:14 | |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 18:14 |
*** PollyZ has quit IRC | 18:17 | |
openstackgerrit | Merged openstack-infra/project-config: Add neutron-dynamic-routing to PROJECT list https://review.openstack.org/340222 | 18:18 |
*** javeriak_ has quit IRC | 18:19 | |
*** tonytan4ever has joined #openstack-infra | 18:20 | |
*** tphummel has joined #openstack-infra | 18:20 | |
*** kushal has quit IRC | 18:22 | |
*** karthikp_ has joined #openstack-infra | 18:23 | |
*** PollyZ has joined #openstack-infra | 18:23 | |
openstackgerrit | Joshua Harlow proposed openstack-infra/irc-meetings: Fix chair for oslo team meetings https://review.openstack.org/341131 | 18:24 |
openstackgerrit | Paul Belanger proposed openstack-infra/project-config: Prefix packaging-deb for reprepro-import URL https://review.openstack.org/341132 | 18:25 |
openstackgerrit | Paul Belanger proposed openstack-infra/project-config: Add reprepro import / release into post pipeline https://review.openstack.org/341133 | 18:25 |
pabelanger | reprepro bits for post queue | 18:25 |
locust | fungi: thank you! | 18:26 |
fungi | locust: you're welcome | 18:26 |
*** mriedem1 has joined #openstack-infra | 18:27 | |
*** Apoorva_ has quit IRC | 18:28 | |
*** Apoorva has joined #openstack-infra | 18:28 | |
*** PollyZ has quit IRC | 18:29 | |
*** mriedem2 has joined #openstack-infra | 18:29 | |
*** mriedem has quit IRC | 18:29 | |
*** eharney has joined #openstack-infra | 18:29 | |
*** locust has quit IRC | 18:31 | |
rcarrillocruz | yolanda: your betamax change just passed neutron func tests | 18:31 |
rcarrillocruz | \o/ | 18:31 |
*** mriedem1 has quit IRC | 18:32 | |
yolanda | yay! | 18:32 |
*** rbrndt has joined #openstack-infra | 18:32 | |
*** sdake_ has quit IRC | 18:32 | |
fungi | as a child i had a betta (fish) named "max" | 18:33 |
*** Sukhdev has joined #openstack-infra | 18:33 | |
*** sdake_ has joined #openstack-infra | 18:33 | |
pleia2 | fungi: haha, that's wonderful | 18:33 |
fungi | his successors were "supermax" and "ed" (for the superbetamax and extended definition formats) | 18:34 |
*** mriedem2 is now known as mriedem | 18:35 | |
pabelanger | pleia2: neat stuff with https://github.com/apache/ and the mirror | 18:35 |
pabelanger | url* | 18:35 |
pleia2 | pabelanger: yeah, I'm thinking our project sync script could just add a prefix in the project description | 18:35 |
*** rdx_matt has joined #openstack-infra | 18:36 | |
pleia2 | and we can update the project description, like drupal and kde have done | 18:36 |
pleia2 | project == openstack (not individual repos) | 18:36 |
pabelanger | pleia2: wonder how we get mirrored from in https://github.com/apache/incubator-pirk | 18:36 |
pabelanger | if that is configurable or not | 18:36 |
pleia2 | pabelanger: ah, I assumed that was the description, but it's not | 18:37 |
* pleia2 pokes around UI | 18:37 | |
*** gouthamr has joined #openstack-infra | 18:38 | |
*** asettle has quit IRC | 18:38 | |
*** shardy is now known as shardy_afk | 18:38 | |
pabelanger | pleia2: I think we'd need to ask github staff: http://stackoverflow.com/questions/11370239/creating-an-official-github-mirror | 18:40 |
clarkb | anyone know who works on blazar? | 18:40 |
pleia2 | it's a mystery | 18:40 |
pleia2 | pabelanger: ah, hm | 18:40 |
clarkb | it has hit AttributeError: 'module' object has no attribute 'PROTOCOL_SSLv3' which I think is a xenial thing due to debuntu removing ssl3 support from python because it is insecure | 18:41 |
clarkb | they should probably not use an insecure protocol | 18:41 |
*** kzaitsev_mb has quit IRC | 18:41 | |
pleia2 | https://help.github.com/articles/about-github-mirrors/ | 18:41 |
*** dizquierdo has quit IRC | 18:42 | |
*** psilvad has quit IRC | 18:42 | |
*** david-lyle has joined #openstack-infra | 18:43 | |
clarkb | pabelanger: I looked into it as an alternative to our push model and iirc you have to aks them and its not autoamtable | 18:43 |
clarkb | which is why we didn't consider changes | 18:43 |
clarkb | s/changes/changing/ | 18:43 |
pabelanger | clarkb: ya | 18:44 |
pleia2 | we can chat more at the meeting, but my main concern is that every day someone submits a change, and we don't give clear indication that it's read-only, or that they should be doing something differently | 18:44 |
clarkb | pleia2: the pull requset closer bot attempts to clarify things | 18:45 |
pleia2 | been thinking about trying to pull some email matching stats to see if any of these people actually follow up after a bot closes their change | 18:45 |
pleia2 | clarkb: people hate that closer bot, they've already put in work | 18:45 |
pleia2 | now they're told to do it differently, it gives them bad feels (and I sympathize) | 18:45 |
fungi | pleia2: also we have used github's recommended convention of providing a CONTRIBUTING.rst file which says these things too | 18:45 |
pleia2 | it's not enough | 18:46 |
clarkb | I don't think we will every be able to do enough as long as we mirror to github | 18:46 |
*** Goneri has quit IRC | 18:46 | |
fungi | i'm afraid nothing will ever be enough to deal with people who don't read and get huffy the first moment they're asked to do something differently | 18:46 |
anteaya | I'm in favour of not mirroring to github | 18:47 |
fungi | same here | 18:47 |
anteaya | I can't please all the people, I accept that about myself | 18:47 |
clarkb | pleia2: fwiw I also checked those apache repos when I was checking out that feature and many of them had pull requests too | 18:47 |
pleia2 | some of the projects I linked in the agenda made it much more obvious than ours, like appending it to the project description that it was read only, etc | 18:47 |
jeblair | pleia2: we could also do what other projects do, and put in a static pull request with contributing info | 18:48 |
fungi | the only real argument i've seen for continuing to mirror to github is that people will search for openstack projects on github because they don't know any better, and if we don't mirror to it then they'll end up finding some joe random outdated fork and assuming that's ours | 18:48 |
pleia2 | then by loading up the project page itself, they can see that they should follow another process without digging into a contributing file | 18:48 |
*** piet has quit IRC | 18:48 | |
pleia2 | jeblair: the pull request template? | 18:48 |
pleia2 | (that's one of my suggestions in the agenda) | 18:48 |
*** PollyZ has joined #openstack-infra | 18:48 | |
clarkb | fungi: many people also use github to collaborate on before pushing to gerrit (whther ot not you want to support that workflow is another question) | 18:49 |
jeblair | pleia2: oh didn't know about that. i meant a pull request that stays open so that people see it when they go to the PR page for a project | 18:49 |
anteaya | I can't even make the people happy who _do_ go to the work of asking themselves what our workflow is | 18:49 |
anteaya | I don't think I can take on the additional numbers of people who can't be bothered to do that | 18:49 |
*** dizquierdo has joined #openstack-infra | 18:49 | |
pleia2 | jeblair: it's new :) https://help.github.com/articles/creating-a-pull-request-template-for-your-repository/ | 18:49 |
*** david-lyle has quit IRC | 18:49 | |
*** adrian_otto has quit IRC | 18:49 | |
*** dizquierdo has quit IRC | 18:49 | |
clarkb | ansible has a template and its annoying because almost never is the stuff in the template useful | 18:49 |
jeblair | pleia2: maybe an 'all of the above' approach, assuming we want to keep mirroring | 18:49 |
*** dizquierdo has joined #openstack-infra | 18:50 | |
pleia2 | jeblair: yeah | 18:50 |
pleia2 | fwiw, I don't like the github mirror more than any of you do, so I'm trying to make the best of it | 18:50 |
anteaya | pleia2: would you be opposed to dropping the github mirror? | 18:50 |
pleia2 | anteaya: no, but people have already mentioned reasons we are keeping it | 18:51 |
fungi | i'm pretty sure we'd have community outrage on our hands if we did | 18:51 |
anteaya | sure we would | 18:51 |
jeblair | thing is, there has always been a github mirror of openstack (even before it used git), and there always will be. the question is -- do we run it. :) | 18:51 |
openstackgerrit | Khai Do proposed openstack/python-jenkins: Add kerberose packages https://review.openstack.org/341143 | 18:51 |
pleia2 | I think a lot of internal corp things we don't see point at github | 18:51 |
fungi | and people would quickly set up hacky pull mirrors into github | 18:51 |
zaro | clarkb: ^ i think that's what you wanted? | 18:51 |
clarkb | zaro: close, you are going to need to list all of the test deps in other-requirements not just the kerberos ones | 18:51 |
clarkb | zaro: python-dev-all etc | 18:52 |
pleia2 | jeblair: I think project integrity requires us to continue in this nasty chore | 18:52 |
* fungi scrambles to do last-minute meeting prep | 18:52 | |
anteaya | jeblair: ah okay, fair | 18:52 |
pleia2 | woo, last agenda item almost finished before the meeting \o/ | 18:52 |
fungi | pleia2: fwiw i'm cool with taking every possible (automated) measure to try and redirect people from github to the correct resources | 18:53 |
*** asettle has joined #openstack-infra | 18:53 | |
*** asettle has quit IRC | 18:53 | |
zaro | does 'tox -e bindep' produce that list? that commands fails for me on my xenail container. | 18:54 |
*** asettle has joined #openstack-infra | 18:54 | |
zaro | clarkb: ^ | 18:54 |
fungi | as long as we accept that whatever we do will never be enough, and that there will always be at least some github zealots who will be incredulous at our insistence to use something else | 18:54 |
pleia2 | fungi: that's helpful, I didn't want to spend time+effort in the weeds if we weren't on the same page, but it seems like we are | 18:54 |
clarkb | zaro: no bindep uses the other-requirements to generate it so you have to proivide it | 18:55 |
*** ddieterly[away] has quit IRC | 18:56 | |
rcarrillocruz | Shrews: we should make a release of shade | 18:56 |
rcarrillocruz | i have stuff broken till we get the fix on pypi : http://logs.openstack.org/23/322823/9/check/gate-ansible-role-cloud-launcher-dsvm-ansible-func-ubuntu-trusty/b7944c7/console.html | 18:56 |
rcarrillocruz | :/ | 18:56 |
*** piet has joined #openstack-infra | 18:57 | |
*** arxcruz has quit IRC | 18:57 | |
*** hurgleburgler has quit IRC | 18:57 | |
jhesketh | Morning | 18:58 |
anteaya | morning jhesketh | 18:58 |
fungi | it's that (weekly infra team meeting) time again! find us in #openstack-meeting for the next hour | 19:00 |
*** ociuhandu has joined #openstack-infra | 19:00 | |
*** daemontool has quit IRC | 19:04 | |
*** bin_ has quit IRC | 19:04 | |
*** ddieterly has joined #openstack-infra | 19:04 | |
*** hurgleburgler has joined #openstack-infra | 19:05 | |
clarkb | zaro: I think you meant to use this channel for bindep? | 19:06 |
clarkb | zaro: look at the other-requirements files for other projects like swift and gear I think | 19:07 |
*** sdake_ has quit IRC | 19:08 | |
Rockyg | clarkb, quickie.... I think Howard is at OpenStack China (Beijing) so may be a deelay in answering | 19:09 |
*** rdx_matt has quit IRC | 19:09 | |
clarkb | Rockyg: thats fine, its not a huge rush. Just want to check in and make sure that there aren't problems pushing upstream etc | 19:09 |
clarkb | Rockyg: I have actually found 10 or so projects liek that and will try to ping all of their owners to see if we can do anything else to help | 19:10 |
Rockyg | kk Thanks. | 19:10 |
clarkb | alsoa few that have jobs setup in zuul that have never been able to pass | 19:10 |
*** nadya has joined #openstack-infra | 19:10 | |
clarkb | no tox.ini, etc | 19:10 |
openstackgerrit | Emilien Macchi proposed openstack-infra/project-config: zuul/tripleo: run gate-tripleo-ci-centos-7-ovb-nonha in tripleo-ci https://review.openstack.org/341144 | 19:11 |
*** nadya has quit IRC | 19:14 | |
*** tidwellr has quit IRC | 19:15 | |
*** PollyZ has quit IRC | 19:17 | |
*** gyee has quit IRC | 19:18 | |
rods | hello, any chance I can get another review from a core for this pr? https://review.openstack.org/#/c/339065/3 | 19:20 |
dhellmann | clarkb : so if I'm reading the scrollback right the releasenotes build issue is with the way the git clone was done? | 19:20 |
clarkb | dhellmann: yes I think so. I haven't gone through and rerun it against all the projects that failed to confirm but I am fairly confident this was it | 19:21 |
clarkb | dhellmann: I think we can ignore that issue for now | 19:21 |
*** hurgleburgler has quit IRC | 19:22 | |
*** flepied has joined #openstack-infra | 19:22 | |
dhellmann | clarkb : cool | 19:23 |
clarkb | dhellmann: still waiting on the last 10 or so projects to go through my test loop (nova and neutron included so may be a while) then I will rerun my script just on releasenotes with that configured properly | 19:23 |
dhellmann | sounds good | 19:23 |
openstackgerrit | Merged openstack-infra/shade: Use keystoneauth.betamax for shade mocks https://review.openstack.org/298647 | 19:24 |
clarkb | dhellmann: the only real issues I have seen so far are an ssl3 related one in blazar (sslv3 was removed from python because its insecure), some potential sqlite3 things and nodepool not liking mysql (but everything else seems fine with it so sqlalchemy/oslo.db are fine I think | 19:24 |
openstackgerrit | Merged openstack-infra/shade: Add support for changing metadata of compute instances https://review.openstack.org/319395 | 19:24 |
clarkb | oh and pbr fails wsgi tests | 19:25 |
zigo | pabelanger: Awesome, thanks a lot. Yes, we should have something like "jessie-mitaka" or "jessie/mitaka", I'd prefer the former, so that we have all in a flat repo. | 19:25 |
*** ajmiller has quit IRC | 19:25 | |
zigo | pabelanger: Also, it needs to come from the built packages, I'm not sure where you got that version 52 from. Is this just from official jessie-backports? | 19:25 |
pabelanger | zigo: okay, I can get a change up for that. It is trivial now | 19:26 |
*** ajmiller has joined #openstack-infra | 19:26 | |
zigo | pabelanger: I'd prefer if we worked on Newton directly though, Mitaka is kind of the past ... :P | 19:26 |
dhellmann | clarkb : weird about the wsgi tests. | 19:26 |
*** sdake_ has joined #openstack-infra | 19:26 | |
*** mikelk has joined #openstack-infra | 19:26 | |
zigo | pabelanger: How does your job work, is it a "post" after the merge of deb-openstack-pkg-tools ? | 19:27 |
pabelanger | zigo: imported from http://tarballs.openstack.org/openstack-pkg-tools/uploads/d2b2b048228e08f77ee0403e5a5f637cc5c6148d/ | 19:27 |
pabelanger | zigo: yes | 19:27 |
pabelanger | zigo: https://review.openstack.org/#/c/341133/ | 19:27 |
zigo | pabelanger: How come it doesn't contain the commit number? | 19:27 |
pabelanger | zigo: what do you mean? | 19:28 |
zigo | pabelanger: Normally, the version should integrate the number of commits before the last tag. | 19:28 |
*** ihrachys has quit IRC | 19:28 | |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Revert "Switch away from the default IP range used by tripleo" https://review.openstack.org/341149 | 19:29 |
zigo | pabelanger: That would be something like this: http://tarballs.openstack.org/packaging-deb/openstack-pkg-tools/uploads/d1ab24a12d2126acccc5c6a56f210ea2b2ae97ea/ | 19:29 |
pabelanger | zigo: that's the version numbering you have done. The stuff I have worked on simply downloaded things from tarball.o.o and imported it into reprepro | 19:29 |
pabelanger | zigo: okay, so I was testing using an old URL | 19:29 |
zigo | :) | 19:29 |
*** sambetts is now known as sambetts|afk | 19:29 | |
pabelanger | moving forward it should be like the URL you listed above | 19:29 |
zigo | pabelanger: We should then agree on what URL scheme to use... | 19:29 |
zigo | pabelanger: Also, I'll need to know what URL to use for the repo. I guess it wont always be rackspace, right? | 19:30 |
openstackgerrit | Merged openstack-infra/tripleo-ci: Revert "Switch away from the default IP range used by tripleo" https://review.openstack.org/341149 | 19:30 |
*** julim has quit IRC | 19:30 | |
pabelanger | zigo: is the URL I am downloading from, you should upload to the same: https://review.openstack.org/#/c/341132/1/jenkins/jobs/packaging-mirror.yaml | 19:30 |
pabelanger | zigo: we can expose that in our nodepool configure_mirror.sh script | 19:30 |
pabelanger | zigo: for example http://git.openstack.org/cgit/openstack-infra/project-config/tree/nodepool/scripts/configure_mirror.sh | 19:31 |
zigo | Yup, URL seems correct. | 19:31 |
*** david-lyle_ has quit IRC | 19:31 | |
pabelanger | zigo: I suggest fixing your jobs on deb-openstack-pkg-tools, they are currently failing | 19:32 |
zigo | pabelanger: Which one? | 19:32 |
pabelanger | and we can finalize the reprepro layout for the codename settings | 19:32 |
pabelanger | zigo: http://logs.openstack.org/20/340520/4/check/pkgdeb-build-openstack-pkg-tools/a5189d1/ | 19:32 |
pabelanger | any new patch fails | 19:32 |
*** esker has joined #openstack-infra | 19:33 | |
zigo | I thought I fixed that ... | 19:33 |
openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul: Call parameter_function before swift parameter https://review.openstack.org/341154 | 19:33 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Configure a different port for httpd ipxe https://review.openstack.org/340339 | 19:34 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 19:34 |
*** ajmiller_ has joined #openstack-infra | 19:34 | |
zigo | pabelanger: The problem seems to be the Unix username. Does "jenkins" still exist as a username? | 19:35 |
*** david-lyle_ has joined #openstack-infra | 19:35 | |
pabelanger | yes | 19:35 |
pabelanger | but it will be changing in the future | 19:35 |
pabelanger | to zuul | 19:35 |
*** ajmiller has quit IRC | 19:38 | |
*** PollyZ has joined #openstack-infra | 19:38 | |
zigo | pabelanger: Ah also, for a reason, patch prosal don't go to the #openstack-pkg channel as they should. I'm not sure what's wrong in what we attempted doing with fungi for that... | 19:38 |
pabelanger | zigo: project-config/gerritbot is what you need to add your project too | 19:39 |
zigo | I believe I did that already, but that's probably wrong. | 19:40 |
* zigo is checking | 19:40 | |
zigo | Yup, it's there... | 19:41 |
zigo | Maybe the bots need a restart? | 19:41 |
pabelanger | zigo: missing mitaka branch | 19:41 |
zigo | Oh ... | 19:41 |
*** hurgleburgler has joined #openstack-infra | 19:42 | |
*** ihrachys has joined #openstack-infra | 19:42 | |
*** PollyZ has quit IRC | 19:43 | |
openstackgerrit | Thomas Goirand proposed openstack-infra/project-config: Add debian/{mitaka,newton} to deb-ospkgtools IRC robot https://review.openstack.org/341157 | 19:44 |
tristanC | jeblair: so it seems like the parameter_functions is called after swift hmac computation, thus to set a custom (swift) LOG_PATH in function.py, we need https://review.openstack.org/#/c/341154 | 19:44 |
*** sarob has joined #openstack-infra | 19:44 | |
*** adrian_otto has joined #openstack-infra | 19:44 | |
zaro | clarkb: thanks for pointer to other-requirements examples but what i'm suggesting is a doc update because what's in the bindep docs doesn't seem like it's enough to direct users to create that file. | 19:44 |
*** _sarob has joined #openstack-infra | 19:45 | |
clarkb | zaro: well its all context dependent on what you need bindep to do | 19:45 |
clarkb | zaro: all bindep is is a list of packages with distro and version markers that when parsed can emit a lsit for your current host | 19:45 |
*** ociuhandu has quit IRC | 19:46 | |
*** ociuhandu has joined #openstack-infra | 19:46 | |
dmsimard | jeblair, tristanC: fyi context: being able to use zuul_ref as the log destination for the artifacts for what we talked about earlier for parent vs child jobs | 19:46 |
zaro | clarkb: at least some references to some of how we use it may help | 19:46 |
*** annegentle has joined #openstack-infra | 19:47 | |
zaro | particularly in the infra manual | 19:47 |
clarkb | ya we could add that there | 19:48 |
zigo | pabelanger: This will fix the issue: https://review.openstack.org/#/c/341155 | 19:48 |
zigo | And should produce another version of openstack-pkg-tools ! :P | 19:48 |
jeblair | dmsimard, tristanC: we should ask jhesketh to review that | 19:49 |
dmsimard | jeblair: ack, I'll add a comment in the review for context. | 19:49 |
*** sarob has quit IRC | 19:49 | |
clarkb | mordred: notmorgan: related to your admin vs not admin slowness on my local test rig for xenial postgres tests are much faster than mysql when running tox -epy27 with keystone | 19:50 |
clarkb | 18seconds vs 4-5seconds | 19:50 |
*** daemontool has joined #openstack-infra | 19:50 | |
pabelanger | AJaeger: ianw: rcarrillocruz: mind helping land 341133 and 341132 | 19:51 |
mordred | clarkb: that's weird and sad | 19:51 |
* jhesketh will add it to the queue | 19:51 | |
openstackgerrit | amrith proposed openstack-infra/irc-meetings: Proposing OpenStack SWG Meeting https://review.openstack.org/338134 | 19:51 |
clarkb | mordred: its possible that mysql is just less happy on my VM than postgres but likely warrants more investigation | 19:51 |
notmorgan | clarkb: good to know. | 19:52 |
*** PollyZ has joined #openstack-infra | 19:52 | |
notmorgan | clarkb: mysql also falls out the gate with horrible defaults | 19:52 |
notmorgan | clarkb: if devstack is doing zero tuning on mysql, you basically have a useless config for nothing more than very limited testing | 19:52 |
clarkb | also I am running serially on one cpu so shouldn' | 19:52 |
clarkb | t be related to tests fighting each other | 19:52 |
*** maishsk has joined #openstack-infra | 19:52 | |
notmorgan | clarkb: if pgsql on the same box is happier, i'd blame mysql's horrible defaults | 19:53 |
*** fawadkhaliq has quit IRC | 19:53 | |
notmorgan | s/box/setup | 19:53 |
clarkb | notmorgan: even 5x the performance cost though? | 19:53 |
notmorgan | you can probably increase buffers and caching and see massive improvements. | 19:53 |
clarkb | its not like we are dealing with massive datasets | 19:53 |
clarkb | its make a user | 19:53 |
notmorgan | you have no idea how bad mysql's defaults are | 19:53 |
clarkb | I have a hard time believing they are that bad | 19:53 |
notmorgan | i wont use them on my dev boxes | 19:53 |
rcarrillocruz | mordred ! | 19:54 |
clarkb | especailly when you consider nodepool uses the same setup just fine | 19:54 |
rcarrillocruz | we need a release of shade | 19:54 |
rcarrillocruz | :-) | 19:54 |
clarkb | this to me screams keystone issue | 19:54 |
clarkb | but what do I know | 19:54 |
notmorgan | nodepool doesn't grind in the db as much | 19:54 |
* stevemar perks | 19:54 | |
notmorgan | having worked with the schemas | 19:54 |
clarkb | notmorgan: neither does keystone in "make a user" | 19:54 |
clarkb | notmorgan: this isn't tempset | 19:54 |
notmorgan | clarkb: it's not "make a user" it's token validation among other things | 19:54 |
clarkb | its the unittest for make a user with mysql abckend | 19:54 |
rcarrillocruz | mordred: https://review.openstack.org/#/c/315713/ merged, solved the gate issues | 19:54 |
notmorgan | which is hit a lot | 19:54 |
*** cody-somerville has quit IRC | 19:55 | |
rcarrillocruz | but we need a release, i have plays failing with fip timeouts :/ | 19:55 |
*** mikelk has quit IRC | 19:55 | |
clarkb | notmorgan: I think you misunderstand me | 19:55 |
clarkb | notmorgan: these are very isolated code paths, its nto a running cloud | 19:55 |
mordred | rcarrillocruz: yah - I saw - I'm poking jlk right now to see if he can verify that it doesn't break him | 19:55 |
rcarrillocruz | ++ | 19:55 |
notmorgan | create a user hits a ton of db grinding to validate things. | 19:55 |
clarkb | and ldap is on the order of miliseconds | 19:55 |
*** karthikp_ has quit IRC | 19:55 | |
clarkb | not sure if thats mocked out though | 19:55 |
notmorgan | ldap is in our testing. | 19:56 |
*** annegentle has quit IRC | 19:56 | |
clarkb | notmorgan: but enough to result in 5x slower? | 19:56 |
notmorgan | clarkb: i could buy 2-3x slowness | 19:56 |
clarkb | I wouldn't be surprised if mysql defaults makes it slower, but its 5x slower | 19:56 |
notmorgan | i've seen it before. | 19:56 |
*** ddieterly is now known as ddieterly[away] | 19:56 | |
notmorgan | it doesn't take a lot to tip over a badly (or completly untuned) mysql. | 19:57 |
*** kzaitsev_mb has joined #openstack-infra | 19:57 | |
*** PollyZ has quit IRC | 19:57 | |
clarkb | I just have a hard time believing a db with no data in it would notice | 19:57 |
clarkb | its not like its going to disk | 19:57 |
clarkb | or needing to rely on buffer tuning | 19:58 |
clarkb | its all in memory because its tiny | 19:58 |
*** ihrachys has quit IRC | 19:58 | |
notmorgan | i can only speak to my experiences. | 19:59 |
*** Apoorva has quit IRC | 19:59 | |
clarkb | yes and I to mine, I am just asking that keystone treat this as a potential issue rather than blame mysql without checking | 19:59 |
notmorgan | i am not a mysql dev/former mysql dev (I defer to mordred, SpamapS, etc) | 19:59 |
*** julim has joined #openstack-infra | 19:59 | |
*** hasharAway is now known as hashar | 20:00 | |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Make ovb testenv more user friendly https://review.openstack.org/337637 | 20:00 |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Update to use quintupleo https://review.openstack.org/336309 | 20:00 |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Deploy full net-iso environment https://review.openstack.org/341161 | 20:00 |
notmorgan | actually, let me make this easier, i really shouldn't speak on keystone's behalf here. | 20:00 |
clarkb | the gate info is likely not clean enough to use in comparison because we run with 8vcpus all fighting each other | 20:00 |
clarkb | but it is easy to reproduce locally with tox -epy27 --concurrency 1 iirc | 20:00 |
clarkb | er -- --concurrency | 20:00 |
* notmorgan has to stop jumping into the keystone discussions | 20:00 | |
*** Apoorva has joined #openstack-infra | 20:00 | |
*** zeih has quit IRC | 20:01 | |
openstackgerrit | Doug Hellmann proposed openstack-infra/project-config: add job to apply tags when based on changes in openstack/releases https://review.openstack.org/341058 | 20:01 |
* notmorgan can't keep all of keystone and the new zuul/infra knowledge swapped in at the same time. | 20:01 | |
*** sdake_ has quit IRC | 20:01 | |
pabelanger | ianw: are we still blocking fedora24 builds with diskimage-builder? Last time I tried I it error'd out on locales | 20:01 |
notmorgan | stevemar: ^ you should look at this | 20:02 |
notmorgan | (the thing clarkb is talking about) | 20:02 |
clarkb | EmilienM: re that logo proposal, regardless of any other reules my undersatnding is any dilution of the openstack logo wouldn't be allowed | 20:02 |
ianw | pabelanger: nope ... should be ready! https://review.openstack.org/#/c/340671/1 | 20:02 |
pabelanger | ianw: ah, nice. I missed that | 20:02 |
EmilienM | clarkb: o | 20:03 |
EmilienM | ok | 20:03 |
ianw | pabelanger: yeah, if we get some nodes i can start transition off f23 | 20:03 |
ianw | pabelanger: also looking at the centos locales thing ... as i think you determined yesterday, guess what, systemd is involved :) | 20:03 |
clarkb | stevemar: basically run the opportunistic mysql and postgres tests with a single threaded test runner and compare the timing between them | 20:04 |
clarkb | stevemar: I am seeing up to 5x slower mysql tests than postgresql | 20:04 |
stevemar | clarkb: i'm still not super clear on what the issue is, partly because i keep checking in half way through | 20:04 |
*** jamesmcarthur has quit IRC | 20:04 | |
pabelanger | ianw: ya, I see some talks about it the other day | 20:04 |
*** pradk_ has joined #openstack-infra | 20:04 | |
clarkb | stevemar: thought I would point it out because I know mordred was looking into another performance related issue with keystnoe admin vs non admin ops | 20:05 |
*** mriedem1 has joined #openstack-infra | 20:05 | |
notmorgan | clarkb: fwiw, i was unable to duplicate the admin vs nonadmin thing eysterday | 20:05 |
stevemar | clarkb: can you file a bug for it? otherwise i'll lose track of it | 20:05 |
*** jcoufal has quit IRC | 20:05 | |
clarkb | stevemar: sure | 20:05 |
*** PollyZ has joined #openstack-infra | 20:05 | |
notmorgan | clarkb: i spent a good deal of time trying to, all examples i could come up with ended up being slow, but about the same slow | 20:05 |
pabelanger | ianw: since you are around: https://review.openstack.org/#/c/341132/ and https://review.openstack.org/#/c/341133 mind a +3? | 20:05 |
pabelanger | ianw: I've tested them already in experimental | 20:06 |
clarkb | notmorgan: interesting | 20:06 |
notmorgan | clarkb: and i was using the same(ish) setup Shrews had. | 20:06 |
*** jamesmcarthur has joined #openstack-infra | 20:06 | |
*** rdx_matt has joined #openstack-infra | 20:06 | |
notmorgan | clarkb: it bothers me that it "magically" fixed or is more subtle. | 20:06 |
notmorgan | clarkb: i just need more concrete "here is how to repo" to do much more. | 20:06 |
zaro | clarkb: so i've found the swift other-requirements.txt but still not sure whether that list overrides or inherits the base set of packages. | 20:07 |
zaro | https://git.openstack.org/cgit/openstack/swift/tree/other-requirements.txt | 20:07 |
notmorgan | or a snapshot/working VM env that i can spin up and poke at with the issue | 20:07 |
zaro | clarkb: i thought it inherits so you only have to add whatever else is missing? | 20:07 |
notmorgan | since i clearly couldn't make an environment that duplicated the issue myself (which just means i didn't have all the data needed to do so, bnot that it isn't an issue) | 20:08 |
*** mriedem has quit IRC | 20:08 | |
*** tonytan4ever has quit IRC | 20:08 | |
*** cody-somerville has joined #openstack-infra | 20:08 | |
zaro | clarkb: i'm still not sure how to currate my own list from that example. | 20:09 |
*** markusry has quit IRC | 20:09 | |
clarkb | stevemar: https://bugs.launchpad.net/keystone/+bug/1602407 | 20:09 |
openstack | Launchpad bug 1602407 in OpenStack Identity (keystone) "MySQL tests significantly slower than PostgreSQL tests" [Undecided,New] | 20:09 |
zigo | pabelanger: Did your job pick-up http://tarballs.openstack.org/packaging-deb/openstack-pkg-tools/uploads/385aacd524eac4c86d93dd58ed1e629637945ff7/ ? | 20:09 |
*** csomerville has joined #openstack-infra | 20:09 | |
clarkb | zaro: it does not inherit, which is intentional we want minimal lists to be clear | 20:09 |
stevemar | clarkb: thank you sir | 20:10 |
*** mriedem has joined #openstack-infra | 20:10 | |
clarkb | zaro: you will need python dev deps. Depends on what python-jenkins needs, will liekly require trial and error | 20:10 |
clarkb | zaro: I would start with what zuul or nodepool use (drop the mysql/postgre stuff) and add in krb5-dev | 20:10 |
*** PollyZ has quit IRC | 20:10 | |
*** karthikp_ has joined #openstack-infra | 20:12 | |
zaro | clarkb: and keep running 'tox -e bindep test' until it passes? | 20:12 |
*** mriedem1 has quit IRC | 20:12 | |
clarkb | no tox -ebindep won't tell you if you have all the packages you need | 20:12 |
*** cody-somerville has quit IRC | 20:12 | |
clarkb | it will only tell you what is missing from the other-requirements.txt list on the current machine iirc | 20:13 |
clarkb | zaro: you have to install the deps from bindep, then run tox -epy27 | 20:13 |
clarkb | if that works then your list is liekly good | 20:13 |
pabelanger | zigo: no, we are still merging to patch | 20:14 |
zaro | ok. will give that a go. again, not sure how other people would figure this out. i've not found it in the documentation. | 20:14 |
pabelanger | zigo: changes are landing now | 20:14 |
fungi | zaro: or propose a change you think is most of the way there and check the job results. other-requirements.txt file addition/removal and changes to that file should all be enturely self-testing in our ci | 20:14 |
fungi | er, entirely | 20:15 |
*** adrian_otto1 has joined #openstack-infra | 20:15 | |
openstackgerrit | Kirill Zaitsev proposed openstack-infra/project-config: Add py27-mitaka tox jobs to murano projects https://review.openstack.org/340954 | 20:15 |
asselin__ | rcarrillocruz, what git format is expected for ansible galaxy? http://paste.openstack.org/show/531014/ | 20:15 |
zaro | clarkb, fungi : thanks, will try to update the docs with this process when i've figured it all out. | 20:16 |
AJaeger | zaro: Quick comment: We have documentation at http://docs.openstack.org/infra/manual/drivers.html#package-requirements and http://docs.openstack.org/infra/bindep/ - if those are confusing, please help updating. | 20:16 |
AJaeger | zaro, thanks! | 20:16 |
fungi | zaro: the way this usually happens from scratch is that you go to run a project's unit tests, it breaks complaining you're missing a c header package to make it possible to build some particular python dependency, so you install that, then you find that it breaks on yet another, and so on. take the list of packages you ended up having to install and put them in an other-requirements.txt file and push | 20:16 |
fungi | up a review for that | 20:16 |
clarkb | asselin__: are you behind a proxy? | 20:16 |
ianw | pliea2: so it has to be a top-of-tree horizon running? i'm wondering if something like RDO can take the pressure off having to maintain the "openstack" portion of the whole thing, by using something more stable than devstack | 20:16 |
clarkb | asselin__: er ya I see it there in the log now | 20:16 |
asselin__ | clarkb, I am | 20:16 |
asselin__ | i had to change away from the git+git protocol | 20:17 |
fungi | ianw: it needs to be master branch tip, right. and then import incomplete translations from zanata on top of that | 20:17 |
rcarrillocruz | asselin__: http://git.openstack.org/cgit/openstack-infra/system-config/tree/roles.yaml | 20:17 |
clarkb | asselin__: our non cgit urls are not web hosted | 20:17 |
clarkb | asselin__: they only do git protocols | 20:17 |
clarkb | asselin__: so wget is doign what I expect | 20:17 |
clarkb | asselin__: try git clone that url | 20:17 |
AJaeger | ianw: it needs a really recent horizon so that the translations that we have in zanata match | 20:17 |
asselin__ | yeah, git clone works fine....but not via ansible galaxy | 20:18 |
fungi | asselin__: the trick is that we have different clone urls and browsing urls. sites/services that assume those will always be identical are broken | 20:18 |
asselin__ | http://paste.openstack.org/show/531015/ | 20:18 |
*** adrian_otto has quit IRC | 20:19 | |
AJaeger | zaro: if you have a change and want feedback, feel free to add me to it or ping me - or ask question via email. I'll drop off again now... | 20:19 |
*** PollyZ has joined #openstack-infra | 20:19 | |
clarkb | asselin__: you can try cloning from the cgit url | 20:19 |
tristanC | jhesketh: nevermind about 341154, I just realise there is already 262112 and 229582... | 20:19 |
clarkb | asselin__: maybe that will make galaxy happy | 20:19 |
fungi | you actually have to do a lot of very hacky browser detection to rewrite browsing requests for your clone urls to something a browser can grok | 20:19 |
*** mrmartin has quit IRC | 20:19 | |
openstackgerrit | Merged openstack-infra/project-config: Prefix packaging-deb for reprepro-import URL https://review.openstack.org/341132 | 20:19 |
fungi | asselin__: clarkb: note that cloning from cgit is extremely terrible from both a performance and reliability standpoint | 20:20 |
asselin__ | sorry, I know i wget it, but that's not really the issue I'm trying to solve....I did that merely to see if there's something that ansible galaxy is also doing | 20:20 |
clarkb | fungi: yup | 20:20 |
asselin__ | by goal is to use a non git url in roles.yaml and ansible galaxy and be happy with | 20:21 |
asselin__ | non git protocol* url | 20:21 |
clarkb | asselin__: yes, try the cgit https url | 20:21 |
clarkb | asselin__: that should at least tell you if galaxy expects a proper web browser for some reason | 20:21 |
zaro | AJaeger: sure will thanks. | 20:22 |
fungi | asselin__: what's in your roles.yaml? | 20:22 |
asselin__ | fungi, http://paste.openstack.org/show/531017/ | 20:23 |
openstackgerrit | Merged openstack-infra/project-config: Add reprepro import / release into post pipeline https://review.openstack.org/341133 | 20:24 |
zaro | fungi: it seems like it's hard to tell what packages are needed and which are appropriate. like i always install 'build-essential' but is that appropriate to put in others-requirement since we want min set? | 20:24 |
pabelanger | zigo: I see an issue with your upload URL, it should be http://tarballs.openstack.org/packaging-deb/deb-openstack-pkg-tools/ not http://tarballs.openstack.org/packaging-deb/openstack-pkg-tools/ because we are using ZUUL_PROJECT | 20:24 |
*** tdasilva has quit IRC | 20:24 | |
asselin__ | clarkb, with cgit url http://paste.openstack.org/show/531018/ | 20:24 |
*** PollyZ has quit IRC | 20:24 | |
asselin__ | not sure if that's right cgit url or not | 20:24 |
*** ddieterly[away] is now known as ddieterly | 20:24 | |
fungi | zaro: yes, that's the crux of the problem bindep is there to solve. hopefully projects adding an other-requirements.txt file reduce the number of devs who have to figure that out to a very small number | 20:25 |
fungi | zaro: but someone still needs to figure it out at least once | 20:25 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Configure a different port for httpd ipxe https://review.openstack.org/340339 | 20:25 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 20:25 |
ianw | pabelanger: i'm not too fussed, we've largely sorted out upload issues but when diagnosing initial devstack runs, it's nice to remove one more variable of bouncing providers | 20:25 |
zigo | pabelanger: will fix | 20:26 |
pabelanger | ianw: Right, I think clarkb mention last time, add all providers out of the box, since things are pretty stable now. But could be wrong | 20:27 |
*** adrian_otto1 has quit IRC | 20:27 | |
clarkb | asselin__: thats the correct url. the error gives you a clue though apparently it wants a tarball | 20:27 |
clarkb | ya I would just add to all providers and cut down on restarting services | 20:27 |
*** Sukhdev has quit IRC | 20:27 | |
*** adrian_otto has joined #openstack-infra | 20:28 | |
*** lucas|afk has quit IRC | 20:28 | |
*** ramishra has quit IRC | 20:29 | |
fungi | asselin__: i suppose we could add branch tarball jobs for those repos and then you could specify https://tarballs.openstack.org/ansible-role-puppet/ansible-role-puppet-master.tar.gz instead | 20:30 |
asselin__ | i'm going to see if I can use github first as a workaround | 20:30 |
*** PollyZ has joined #openstack-infra | 20:31 | |
*** ramishra has joined #openstack-infra | 20:31 | |
openstackgerrit | Khai Do proposed openstack-infra/system-config: Install the Gerrit its-storyboard plugin https://review.openstack.org/340605 | 20:31 |
clarkb | or maybe there us a way to tell galaxy the url is a repo | 20:31 |
clarkb | rather than a tarball | 20:31 |
fungi | clarkb: oh, i think mordred might have submitted a pull request for this actually | 20:31 |
fungi | or i might be confusing it with go get | 20:32 |
*** deadnull has quit IRC | 20:33 | |
*** kgiusti has left #openstack-infra | 20:33 | |
rcarrillocruz | anteaya: nice, didn't know the beaver was canada's national animal :D | 20:34 |
*** lucasagomes has joined #openstack-infra | 20:35 | |
fungi | clarkb: asselin__: yeah, it looks like there are some additional parameters you can specify | 20:35 |
fungi | things like "path" | 20:35 |
mgagne | it isn't the moose :D | 20:35 |
asselin__ | btw, this url work: - src: https://github.com/openstack-infra/ansible-role-puppet | 20:35 |
*** csomerville has quit IRC | 20:35 | |
fungi | also it looks like convention if you want to have it use git is to specify git+https:// otherwise it thinks you're referring to a file | 20:35 |
fungi | github may do fancy rewrites on the backend to make that just work | 20:36 |
asselin__ | i found this option too: scm: git | 20:37 |
jeblair | jhesketh, mtreinish, fungi: i don't understand where prometheus is coming from or why it's in 329115 | 20:37 |
openstackgerrit | Roland Hochmuth proposed openstack-infra/project-config: Add openstack/monasca-grafana-datasource project https://review.openstack.org/341176 | 20:38 |
jeblair | jhesketh, mtreinish, fungi: i think the mqtt thing is a great idea. i don't see that it has anything to do with prometheus, and i'm concerned about +1ing the spec without understanding what is intended there. | 20:38 |
fungi | jeblair: jhesketh seemed to _really_ want someone to mention that a monitoring system could maybe make use of this in the future, enough to -1 the spec | 20:38 |
fungi | i don't personally see that happening, but *shrug* | 20:39 |
openstackgerrit | C.J. Collier proposed openstack-infra/gerritbot: Improve gerritbot error handling https://review.openstack.org/287056 | 20:39 |
asselin__ | clarkb, fungi rcarrillocruz fyi I got it working by adding scm: git: http://paste.openstack.org/show/531019/ | 20:39 |
rcarrillocruz | nice! | 20:39 |
pabelanger | zigo: okay, once you fix the URL, it should publish the package | 20:39 |
jeblair | fungi, jhesketh: that feels like a forced scope expansion to me. this is an open ended spec that says "hey people can do whatever with this". i don't see the point of mentioning prometheus unless the goal is to say "we agreed in a spec to implement prometheus" which, if we want to do, we need to have a conversation about | 20:39 |
*** zz_ja is now known as zz_zz_ja | 20:40 | |
pabelanger | zigo: tomorrow we'll address the codename issue | 20:40 |
mordred | fungi: what did I do? | 20:40 |
fungi | mordred: something else i misremembered | 20:40 |
*** ddieterly is now known as ddieterly[away] | 20:40 | |
mordred | asselin__, clarkb: oh - galaxy and roles and infra? | 20:40 |
fungi | mordred: updating go tools to know how to handle our git server urls | 20:40 |
mordred | yes. that has been merged | 20:40 |
jeblair | fungi, jhesketh: i would be *far* more comfortable if someone wants to stand up prometheus, that there be a spec for that | 20:40 |
*** rdx_matt has quit IRC | 20:41 | |
mordred | asselin__: yeah -for now galaxy only indexes roles from github. I've chatted with the galaxy team about being able to register non-github git locations but it has not hit their todo list yet (understaffed) | 20:42 |
fungi | jeblair: i'm happy to take it back out. as i said in my review comment i don't want to plan for this to be relied on for production purposes until we know it's worth running in the long-term, which will take a while to figure out | 20:42 |
*** ccamacho has quit IRC | 20:42 | |
*** csomerville has joined #openstack-infra | 20:42 | |
mordred | however, because we changed names to ansible-role-*foo* form, it's fairly easy to point galaxy at our github proxy orgs and have it index all of the ansible-role-* repos it finds I've been meaning to write something up about that | 20:42 |
asselin__ | mordred, actually it works fine using the github url. For the openstack url I needed to add the scm: git option. | 20:43 |
fungi | jeblair: i also have a feeling some of jhesketh's concern might have stemmed from overzealous scope wording in the original patchset, which i softened up quite a bit when i did my pass. so maybe it's addressed even without mentioning monitoring systems and failure prediction | 20:44 |
*** shardy_afk is now known as shardy | 20:44 | |
fungi | i'll pull that sentence back out and see where we end up | 20:44 |
jeblair | fungi: possibly so -- maybe it's "keeping track of what each service is doing" that's the problem | 20:44 |
fungi | jeblair: ooh, good eye. i'll reword that bit real quickly too | 20:45 |
jeblair | that's a pretty ambiguous phrase -- while mqtt certainly lets you see what things are doing (it's an event stream!), it may not actually help "keep track" of things (it's ephemeral!) | 20:45 |
*** david-lyle_ has quit IRC | 20:45 | |
*** mdrabe_ has joined #openstack-infra | 20:46 | |
*** mdrabe has quit IRC | 20:46 | |
*** karthikp_ has quit IRC | 20:46 | |
asselin__ | fungi: this also works: - src: git+https://git.openstack.org/openstack-infra/ansible-role-puppet | 20:46 |
*** chem has quit IRC | 20:46 | |
fungi | jeblair: yep, heartily agreed | 20:46 |
mordred | asselin__: ah - nod. I misunderstood what you were trying to do | 20:46 |
*** karthikp_ has joined #openstack-infra | 20:47 | |
mordred | asselin__: I would like for you to just be able to reference openstack-infra.puppet | 20:47 |
*** priteau has joined #openstack-infra | 20:47 | |
mordred | but that's a little bit more work | 20:47 |
*** akshai has quit IRC | 20:48 | |
asselin__ | mordred, basically working around git protocol blocks :) | 20:48 |
*** esker has quit IRC | 20:48 | |
*** gyee has joined #openstack-infra | 20:48 | |
*** tdasilva has joined #openstack-infra | 20:48 | |
mordred | :) | 20:48 |
*** david-lyle_ has joined #openstack-infra | 20:49 | |
asselin__ | but I like the scm option. It's clear what it's trying to do. | 20:49 |
*** markvoelker has quit IRC | 20:49 | |
sridhar_ram | infra-team: I'm trying understand diff reason why a patchset wouldn't merge after a +A. Looking at https://review.openstack.org/#/c/332097/ I'm wondering how I would go about finding the reason ? Does "Needs Verified" indicate this patchset needs a newer Jenkins run ? | 20:49 |
sridhar_ram | *diff reasons | 20:50 |
mgagne | sridhar_ram: make sure parent change is merged: https://review.openstack.org/#/c/331304/ | 20:50 |
mgagne | sridhar_ram: I think you asked about the same last week: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2016-07-08.log.html#t2016-07-08T16:27:47 | 20:51 |
openstackgerrit | Jeremy Stanley proposed openstack-infra/infra-specs: Firehose: A unified message bus for Infra services https://review.openstack.org/329115 | 20:51 |
sridhar_ram | mgagne: ouch, same reason again.. how did you find that out ? | 20:51 |
fungi | jeblair: jhesketh: ^ | 20:51 |
mgagne | it's the first thing I check because it's often/always the same reason =) | 20:51 |
*** _ari_ has quit IRC | 20:51 | |
ianw | pleia2 / AJaeger : if you moved the zanata pull into a normal job and did a "sleep ((20 * 60))" (assuming we figured out http access) would that be suitable, i.e. enough time to log-in and check your translations? | 20:51 |
sridhar_ram | mgagne: yes, i did :( | 20:51 |
jeblair | fungi: lovely! | 20:52 |
fungi | pabelanger: i've (minorly) updated 329115 | 20:52 |
pleia2 | ianw: the main problem is that most translations don't end up in jobs at all, if the translations aren't 75% complete we don't sync them into git | 20:52 |
pleia2 | ianw: but we also just need a casual system where folks can log in whenever and check the status, not a time limit on availability | 20:53 |
sridhar_ram | mgagne: i think, now this will be itched in my mind 8) thanks a bunch | 20:53 |
mgagne | =) | 20:53 |
ianw | pliea2: i just mean a job you can run ad-hoc when you know you want to check things out | 20:53 |
pleia2 | ianw: well, it's not me :) I only speak English, it's the translations folks who mostly aren't developers and need something simple | 20:54 |
fungi | ianw: also they're not going to be thrilled when they have to wait 30 minutes to see the resulting site | 20:54 |
fungi | (and that's on a good day when we're not resource-starved) | 20:55 |
ianw | i just think that at any given time devstack is going to be unstable, i can't see how we can do bleeding edge but not bleed | 20:55 |
pleia2 | yeah, many of the translators are unpaid, they do this in free time here and there | 20:55 |
ianw | but we're quite good at spinning up a fresh vm and deploying onto that at any given time | 20:55 |
pleia2 | ianw: instead of a human triggering a job, if we could just do it daily automatically it could work | 20:56 |
*** gouthamr has quit IRC | 20:56 | |
pabelanger | fungi: ack | 20:56 |
*** piet has quit IRC | 20:56 | |
fungi | pleia2: ianw: it's also entirely valid to admit defeat on this. demo'ing current un-merged translation state live over the internet in an automated fashion may simply not be something the inherent instability (due to the pace of openstack developent) lends itself to. it will be disappointing to quite a few people, but reality is what it is sometimes | 20:57 |
*** shashank_hegde has quit IRC | 20:57 | |
*** ccamacho has joined #openstack-infra | 20:58 | |
*** stewie925 has joined #openstack-infra | 20:58 | |
fungi | continuous deployment of openstack from master is nothing to sneeze at | 20:58 |
wznoinsk_ | fungi, hi, apologies I missed the infra meeting | 20:58 |
*** wznoinsk_ is now known as wznoinsk | 20:58 | |
mordred | fungi: yah. that's what I was thinking earlier | 20:58 |
fungi | lots of companies have whole teams of devs and operators whose job it is to make that possible | 20:58 |
mordred | fungi: itsounds like a description of a CD-from-master openstack | 20:59 |
fungi | and even they can't always do it reliably | 20:59 |
mordred | yup. paste files. rootwrap files. ... stuff changes | 20:59 |
pleia2 | glad it's not just me, this project has been frustrating and depressing | 20:59 |
ianw | fungi / pleia2 : it just seems like https://git.openstack.org/cgit/openstack-infra/puppet-translation_checksite/tree/templates/zanata-sync.sh.erb could be done quite easily in a job that would be quite reliable. i know you'd have to wait a bit for output, and we could only hold a node for a little while, but it might be enough? | 21:00 |
mordred | up until recently I had a team of almost 8 working full-time on that - it's non-trivial | 21:00 |
*** rfolco has quit IRC | 21:00 | |
pleia2 | I thought it would be made easier by being short-lived and respin every week without preserving data, but even that :( | 21:00 |
*** markvoelker has joined #openstack-infra | 21:00 | |
fungi | ianw: hacky solution, we could set the job timeout to 24 hours for it. then we just need some sort of mechanism to update dns automatically or manage a redirect maybe | 21:01 |
*** shashank_hegde has joined #openstack-infra | 21:01 | |
mordred | pleia2: that was the original remit of trystack 4-5 years ago | 21:01 |
* fungi is mostly joking about the 24-hour job timeout | 21:01 | |
*** zeih has joined #openstack-infra | 21:01 | |
*** jamesmcarthur has quit IRC | 21:02 | |
*** jamesmcarthur has joined #openstack-infra | 21:02 | |
fungi | a mechanism to refresh the deployment on a new system and then only update the redirect site if that succeeds is probably a clean-ish solution | 21:02 |
*** claudiub has quit IRC | 21:02 | |
jeblair | mordred: i think it's worth noting that this system doesn't need to do much; it's an all-in-one, and doesn't need to support any real capacity or deal with real networks, etc... | 21:02 |
*** ddieterly[away] is now known as ddieterly | 21:03 | |
jeblair | mordred: so i take your point that deploying openstack is hard -- just that this problem is more constrained than those examples | 21:03 |
*** sdake has quit IRC | 21:03 | |
*** akshai has joined #openstack-infra | 21:03 | |
pleia2 | fungi: yeah | 21:03 |
fungi | if we were at the point with our launch-node automation we've been dreaming about where we can launch new servers independently of ci jobs, this would be a relatively minor addition on top | 21:03 |
clarkb | stevemar: keystone also has args already parsed errors building docs | 21:04 |
fungi | managing a web redirect instead of updating dns also makes it more tractable | 21:04 |
jeblair | fungi: this is a nice problem for zuulv3, bet we could probably approximate what you just suggested in v2 | 21:04 |
fungi | something we can host in a trivial vhost on static.o.o for example | 21:04 |
*** david-lyle has joined #openstack-infra | 21:04 | |
*** hurgleburgler has quit IRC | 21:04 | |
jeblair | fungi: (a zuul job to deploy with a child job to update the redirect if it succeeds) | 21:04 |
pleia2 | jeblair: that would be nice | 21:05 |
*** xenogear has quit IRC | 21:05 | |
fungi | jeblair: the trick is, if the job succeeds, we also want to keep the worker around for some long but undefined period, and only reclaim it if the next job run succeeds | 21:05 |
stevemar | clarkb: yay, file a bug for that one too? should be easier to fix than your other :) | 21:06 |
fungi | jeblair: so still dynamically launched/deleted nodes, but with different deletion conditions than nodepool currently handles | 21:06 |
jeblair | fungi, pleia2: if whatever deployment system you choose has a way of resetting to zero at the start, then you can do it on static nodes (i could see doing a/b switching between static nodes being pretty easy) | 21:07 |
*** piet has joined #openstack-infra | 21:07 | |
ianw | i feel like this might be tending towards devstack running in dib ... | 21:07 |
fungi | yep, a/b alternating would be an option as long as we continue deploying to b until b succeeds and don't alternate when there's a job failure | 21:07 |
*** hurgleburgler has joined #openstack-infra | 21:07 | |
clarkb | stevemar: https://bugs.launchpad.net/keystone/+bug/1602422 | 21:09 |
openstack | Launchpad bug 1602422 in OpenStack Identity (keystone) "Many tracebacks building keystone docs" [Undecided,New] | 21:09 |
*** jamesmcarthur has quit IRC | 21:09 | |
pleia2 | eep, lunch is here | 21:09 |
*** trown is now known as trown|outtypewww | 21:09 | |
fungi | i need to help with dinner myself | 21:09 |
*** jamesmcarthur has joined #openstack-infra | 21:09 | |
*** shashank_hegde has quit IRC | 21:09 | |
stevemar | clarkb: building them now myself | 21:09 |
stevemar | oh yeah, i see them now | 21:10 |
*** _sarob has quit IRC | 21:10 | |
clarkb | stevemar: ya I realized there was a whoel bunch in addition to the duplicate arg exception so didn't try to capture all of them and just linked the log instead | 21:11 |
clarkb | (which will be deletedin 4months but meh) | 21:11 |
stevemar | clarkb: i think we can fix it up in 4 months | 21:12 |
*** jamesmcarthur has quit IRC | 21:13 | |
* ianw gets breakfast to complete the infra meal cycle | 21:13 | |
*** jamesmcarthur has joined #openstack-infra | 21:13 | |
wznoinsk | fungi we plan, given the approval, to have a joint talk (Intel/Mellanox) about - 3rdparty Openstack CI - testing NFV and SR-IOV | 21:13 |
wznoinsk | abstract - https://etherpad.openstack.org/p/3rdparty_Openstack_CI_-_testing_NFV_and_SR-IOV | 21:14 |
jeblair | now i want afternoon tea | 21:14 |
clarkb | stevemar: also not trying to pick on keystone its just what was goign through my giant loop of run all the tests :) | 21:14 |
stevemar | clarkb: oh np | 21:14 |
stevemar | i understand | 21:14 |
wznoinsk | fungi, I wanted to get os infra ppl to comment on it, give some hints on what the community is looking for at the moment so we could include it in the presentation | 21:14 |
stevemar | ya find one issue and then another 4 come up | 21:14 |
clarkb | stevemar: I am filtering through the fails to figure out which are related to running on xenial and which are problems in the projects before we do the switch to testing on xenial | 21:15 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 21:15 |
stevemar | clarkb: makes sense, i appreciate all the feedback | 21:15 |
*** thorst has quit IRC | 21:16 | |
*** thorst has joined #openstack-infra | 21:17 | |
openstackgerrit | James E. Blair proposed openstack-infra/project-config: Add puppet-ipsilon https://review.openstack.org/341192 | 21:17 |
*** thorst_ has joined #openstack-infra | 21:18 | |
jeblair | fungi: https://review.openstack.org/341193 Add puppet-ipsilon to infra | 21:19 |
*** thorst__ has joined #openstack-infra | 21:20 | |
clarkb | mordred: when you are actually awake do yopu think you will have time to look at pbr wsgi tests failing on xenial? | 21:20 |
clarkb | lifeless: Nakato ^ that may interest you as well | 21:20 |
*** tonytan4ever has joined #openstack-infra | 21:20 | |
*** rcernin has joined #openstack-infra | 21:21 | |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Configure a different port for httpd ipxe https://review.openstack.org/340339 | 21:21 |
openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add mulitnode CI job support to tripleo-ci https://review.openstack.org/324777 | 21:21 |
mordred | clarkb: maybe? I'm about to get on a plane - not sure what the day looks like for hacking | 21:21 |
*** maishsk_ has joined #openstack-infra | 21:21 | |
*** thorst has quit IRC | 21:21 | |
clarkb | ok | 21:22 |
*** esker has joined #openstack-infra | 21:22 | |
mordred | clarkb: is there a link anywhere? or just need to get a xenial vm and run them? | 21:22 |
*** zeih has quit IRC | 21:22 | |
*** thorst_ has quit IRC | 21:23 | |
clarkb | mordred: ya just need xenial and run them | 21:23 |
clarkb | mordred: I can paste the traceback I got locally | 21:23 |
mordred | nah - that won't help debugging | 21:24 |
mordred | I'll find some xenial when I get a chance | 21:24 |
*** thorst__ has quit IRC | 21:24 | |
*** maishsk has quit IRC | 21:24 | |
*** maishsk_ is now known as maishsk | 21:24 | |
clarkb | mordred: http://paste.openstack.org/show/531023/ | 21:24 |
clarkb | well its there anyways | 21:24 |
mordred | cool. thanks | 21:24 |
clarkb | its in the subprocess stdout pip reading | 21:24 |
clarkb | we expect a string to be written and get empty string | 21:24 |
clarkb | I haven't had time to go install and run that helper command outside the test suite to see if it does something funny | 21:25 |
clarkb | notmorgan: and I think it was you that had mentioned addressing mysql on xenial for nodepool? | 21:26 |
clarkb | notmorgan: is that code up yet? | 21:26 |
*** tonytan4ever has quit IRC | 21:26 | |
*** esker has quit IRC | 21:27 | |
*** PollyZ has quit IRC | 21:27 | |
*** esker has joined #openstack-infra | 21:28 | |
*** ldnunes has quit IRC | 21:31 | |
*** jamesmcarthur has quit IRC | 21:31 | |
*** csomerville has quit IRC | 21:31 | |
*** jamesmcarthur has joined #openstack-infra | 21:31 | |
*** csomerville has joined #openstack-infra | 21:31 | |
*** david-lyle has quit IRC | 21:31 | |
*** priteau has quit IRC | 21:32 | |
*** PollyZ has joined #openstack-infra | 21:32 | |
*** priteau has joined #openstack-infra | 21:33 | |
*** asettle has quit IRC | 21:33 | |
*** esker has quit IRC | 21:33 | |
*** ramishra has quit IRC | 21:35 | |
*** hashar has quit IRC | 21:36 | |
*** dprince has quit IRC | 21:36 | |
*** ramishra has joined #openstack-infra | 21:36 | |
*** jamesmcarthur has quit IRC | 21:36 | |
*** rdx_matt has joined #openstack-infra | 21:37 | |
*** shardy has quit IRC | 21:38 | |
*** jamesmcarthur has joined #openstack-infra | 21:39 | |
*** jamesmcarthur has quit IRC | 21:39 | |
*** EricGonczer_ has quit IRC | 21:42 | |
openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: WIP: ZooKeeper API https://review.openstack.org/340658 | 21:43 |
notmorgan | clarkb: it is | 21:44 |
edmondsw | mordred, the link in the Q release naming email I just got doesn't work, indicates the key is invalid... known issue? | 21:44 |
clarkb | notmorgan: cool I will have to dig it up and double check it makes xenial happy that review | 21:45 |
notmorgan | clarkb: https://review.openstack.org/#/c/338619/ | 21:45 |
notmorgan | clarkb: considering i wrote and test it on xenial here, i *hope* it works ;) | 21:45 |
notmorgan | and makes things happy for you | 21:45 |
clarkb | edmondsw: yes see the thraed on the mailing list | 21:45 |
edmondsw | ah, I see it now... tx | 21:46 |
*** amitgandhinz has quit IRC | 21:47 | |
clarkb | TIL about https://git.openstack.org/cgit/openstack/kuryr-libnetwork/tree/tools/tox_install.sh | 21:47 |
*** adrian_otto has quit IRC | 21:50 | |
*** abregman|mtg has quit IRC | 21:50 | |
*** yaume_ has quit IRC | 21:50 | |
clarkb | jeblair: I think http://paste.openstack.org/show/531030/ is new git interacting with GitPython in ways it doesn't expect | 21:50 |
clarkb | jeblair: what is odd is that the other tox targets for that repo which use the tox install helper script don't hit the same issue | 21:51 |
*** yaume has joined #openstack-infra | 21:51 | |
clarkb | I wonder if itis a race of some sort | 21:51 |
openstackgerrit | Khai Do proposed openstack/python-jenkins: Setup required packages with bindep https://review.openstack.org/341143 | 21:53 |
zaro | clarkb: ^ | 21:53 |
*** clenimar_ has joined #openstack-infra | 21:54 | |
jeblair | clarkb: hrm, the py35 cloner tests on zuul are all failing for some other reason; i don't think they're even getting that far | 21:54 |
clarkb | jeblair: oh right py35 should exercise that too (this ran on python27 but underlying git should be the same) | 21:55 |
*** burgerk has quit IRC | 21:55 | |
*** tphummel has quit IRC | 21:55 | |
jeblair | clarkb: yeah, was just hoping for some test results to easy narrow in on that, but no such luck | 21:55 |
jeblair | clarkb: i'll grab a xenial node and poke at it | 21:55 |
clarkb | jeblair: thanks | 21:56 |
openstackgerrit | Marc Abramowitz proposed openstack-infra/jenkins-job-builder: Refactor delete_jobs https://review.openstack.org/88646 | 21:56 |
*** tphummel has joined #openstack-infra | 21:57 | |
*** dtardivel has quit IRC | 21:57 | |
*** Sukhdev has joined #openstack-infra | 21:57 | |
*** markvoelker has quit IRC | 21:59 | |
*** akshai has quit IRC | 22:00 | |
*** markvoelker has joined #openstack-infra | 22:04 | |
Shrews | grr... testtools getUniqueInteger() does not appear to be giving me a unique integer per test | 22:05 |
clarkb | Shrews: its likely only unique per process | 22:05 |
Shrews | that'd likely be it | 22:06 |
* Shrews shakes fist at concurrency demons | 22:06 | |
*** esberglu has quit IRC | 22:06 | |
*** clenimar_ has quit IRC | 22:06 | |
clarkb | Shrews: if you only need one unique integer per process use the pids | 22:07 |
clarkb | if you need mulitple unique integers per process use random uuid or something | 22:07 |
Shrews | good idear | 22:07 |
clarkb | now to figure out why poppy can't import things on xenial | 22:07 |
clarkb | I think its a circular import but I just don't see it | 22:08 |
*** mriedem has quit IRC | 22:08 | |
*** priteau has quit IRC | 22:09 | |
*** priteau has joined #openstack-infra | 22:10 | |
clarkb | or maybe its shadowing something | 22:10 |
*** netsin has quit IRC | 22:10 | |
*** rdx_matt has quit IRC | 22:11 | |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: TEST: Run undercloud idempotency test on all jobs https://review.openstack.org/336112 | 22:11 |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Add undercloud idempotency test https://review.openstack.org/279218 | 22:11 |
*** priteau has quit IRC | 22:13 | |
clarkb | nova unittests pass on xenial | 22:13 |
openstackgerrit | Anand Desai proposed openstack-infra/glean: Disable live check on configured interfaces. Change-Id: I566b10393c09f73dec9f3ecd1f03f8a626591354 Signed-off-by: Anand Desai <anandd@us.ibm.com> https://review.openstack.org/340585 | 22:14 |
*** cbader has quit IRC | 22:14 | |
jeblair | | 2218349 | rax-dfw | None | centos-7 | zuul | zl03 | centos-7-rax-dfw-2218349 | centos-7-rax-dfw-2218349 | 20058041-ea30-4c65-aeab-2fdc77505165 | 23.253.94.57 | hold | 07:10:51:13 | None | 22:15 |
jeblair | infra-root: ^ is that needed? (it was held with no reason) | 22:15 |
clarkb | I do not know what it is held for | 22:16 |
clarkb | fine by me to remove | 22:16 |
jeblair | (also, maybe we should make --reason non-optional) | 22:16 |
clarkb | mordred: I think jd__'s pifpaf thing may possibly have the same problem as pbr's wsgi tests. ERROR: pifpaf: Program did not print: `Available at http://' | 22:16 |
clarkb | jeblair: that seems like a good idea | 22:16 |
clarkb | mordred: could be something up with reading pipes on xenial? I have no idea but about to check pifpafs code to see if it does the same thing as pbr's tests | 22:17 |
*** karthikp_ has quit IRC | 22:17 | |
*** zeih has joined #openstack-infra | 22:19 | |
clarkb | mordred: https://github.com/jd/pifpaf/blob/e2f255c93bb58748e1a1bf3665d077d91759414e/pifpaf/drivers/__init__.py#L122-L131 looks pretty similar, if not line would match empty string | 22:19 |
jeblair | clarkb: http://paste.openstack.org/show/531035/ | 22:19 |
*** clenimar_ has joined #openstack-infra | 22:19 | |
clarkb | jeblair: ya I think it must be a race somehow | 22:20 |
clarkb | jeblair: where git emits string GitPython isn't ready for. Its possible that this is an issue on trusty even we just never see it | 22:20 |
*** shashank_hegde has joined #openstack-infra | 22:21 | |
*** david-lyle_ is now known as david-lyle | 22:22 | |
jeblair | clarkb: ah, there are some hits for "Could not parse git-remote prune result" in logstash.... | 22:23 |
clarkb | yay logstash | 22:23 |
*** xyang1 has quit IRC | 22:23 | |
clarkb | jeblair: any chance those hits happened on not trusty? | 22:24 |
jeblair | clarkb: appears on trusty, precise, xenial, centos7, f23 | 22:24 |
clarkb | yay?! | 22:24 |
clarkb | ok at least it isn't a regression on xenial | 22:24 |
*** eeiden has joined #openstack-infra | 22:24 | |
*** zeih has quit IRC | 22:25 | |
*** thorst has joined #openstack-infra | 22:25 | |
jeblair | ya, highly clustered by time -- all the events in logstash are 07-05 21:xx -- 07-06 01:xx | 22:25 |
jeblair | i think 21 instances of it | 22:26 |
jeblair | almost all infra puppet jobs | 22:27 |
jeblair | s/almost// | 22:27 |
*** tpsilva has quit IRC | 22:28 | |
jhesketh | fungi, jeblair: sorry I missed the conversation earlier re firehose.. I haven't read the new patch yet, but I think my comment may have been mis-understood. It's not that I want the spec to cover monitoring, but I want the spec to limit itself in scope to not include monitoring (which I worded poorly by saying it should mention monitoring outside of the system and gave prometheus as an example) | 22:28 |
jhesketh | fungi, jeblair: the issue (from memory) was there were parts of the spec that implied the state of infra could be determined from the firehose. My comment was that I didn't think that was the point or the right tool. | 22:28 |
*** thorst_ has joined #openstack-infra | 22:29 | |
*** Sukhdev has quit IRC | 22:29 | |
*** larainema has quit IRC | 22:29 | |
*** ddieterly has quit IRC | 22:29 | |
*** thorst has quit IRC | 22:30 | |
jeblair | jhesketh: cool, i think we're on the same page then -- i think fungi and i figured that was probably the issue. fungi pushed up a new patchset that hopefully addresses that (no mention of prometheus, but also removes the suggestion that this could be used to "keep track of state") | 22:30 |
*** jlk has joined #openstack-infra | 22:31 | |
clarkb | stevemar: still about? | 22:31 |
clarkb | stevemar: have question about some oddness I am seeing with openstackclient tests | 22:31 |
clarkb | stevemar: basically hitting AttributeError: 'module' object has no attribute 'select_auth_plugin' and sure enough if I look at https://git.openstack.org/cgit/openstack/osc-lib/tree/osc_lib/api/auth.py select_auth_plugin isn't there but the python27 tests pass in zuul | 22:32 |
clarkb | stevemar: http://paste.openstack.org/show/531037/ is a full traceback | 22:32 |
*** thorst_ has quit IRC | 22:34 | |
*** rcernin has quit IRC | 22:35 | |
*** ramishra has quit IRC | 22:35 | |
*** zul has quit IRC | 22:36 | |
*** ramishra has joined #openstack-infra | 22:36 | |
*** zul has joined #openstack-infra | 22:37 | |
*** tonytan4ever has joined #openstack-infra | 22:37 | |
*** somerville32 has joined #openstack-infra | 22:39 | |
*** somerville32 has joined #openstack-infra | 22:39 | |
*** gordc has quit IRC | 22:40 | |
*** rossella_s has quit IRC | 22:40 | |
*** ajmiller_ has quit IRC | 22:40 | |
clarkb | stevemar: actually I think I may see hat is going on, will patch if this is the issue | 22:40 |
*** rossella_s has joined #openstack-infra | 22:41 | |
*** tonytan4ever has quit IRC | 22:42 | |
openstackgerrit | Merged openstack-infra/zuul: Use the new launch client in zuul server https://review.openstack.org/304132 | 22:42 |
clarkb | oh looks like its already patched upstream | 22:42 |
clarkb | also I think this means anyone installing openstackclient right now without constraints is broken | 22:42 |
*** csomerville has quit IRC | 22:43 | |
clarkb | ah only from source | 22:43 |
*** csomerville has joined #openstack-infra | 22:43 | |
jhesketh | jeblair: cool, will look shortly :-) | 22:44 |
*** weshay has quit IRC | 22:45 | |
*** Julien-zte has joined #openstack-infra | 22:46 | |
jeblair | jhesketh: also, i'm reviewing your v3 backlog :) | 22:47 |
*** somerville32 has quit IRC | 22:47 | |
jhesketh | jeblair: yay, although I'm not even sure what I have in my v3 backlog :-p | 22:48 |
openstackgerrit | Marc Abramowitz proposed openstack-infra/jenkins-job-builder: Refactor delete_jobs https://review.openstack.org/88646 | 22:48 |
jhesketh | jeblair: I do have a bunch of v2 things that should possibly be considered though | 22:48 |
clarkb | oh hey trove still uses proboscis | 22:51 |
clarkb | this fail might be my first one that I can legitly tie back to xenial | 22:51 |
*** rbrndt has quit IRC | 22:51 | |
clarkb | mysql emits a warning when you passwd on the cli now | 22:51 |
*** dimtruck is now known as zz_dimtruck | 22:52 | |
*** mriedem has joined #openstack-infra | 22:53 | |
jeblair | Shrews: can you use the topic "nodepool-zk" for the zookeeper stuff? http://specs.openstack.org/openstack-infra/infra-specs/specs/nodepool-zookeeper-workers.html#gerrit-topic | 22:54 |
*** pradk has quit IRC | 22:54 | |
anteaya | rcarrillocruz: https://en.wikipedia.org/wiki/Canada#Symbols glad to share | 22:56 |
openstackgerrit | greghaynes proposed openstack/diskimage-builder: WIP: Move hook generation in to python https://review.openstack.org/271139 | 22:56 |
dtroyer | clarkb: I think https://review.openstack.org/#/c/339077/ will fix OSC | 22:57 |
clarkb | dtroyer: ya I figured it out | 22:57 |
dtroyer | also https://review.openstack.org/#/c/329189/ | 22:57 |
openstackgerrit | James E. Blair proposed openstack-infra/zuul: Add dynamic reconfiguration https://review.openstack.org/340597 | 22:57 |
anteaya | rcarrillocruz: I see spain's national animal is the bull, makes sense | 22:57 |
*** csomerville has quit IRC | 23:00 | |
fungi | jeblair: that held node is not one of mine, at least as far as i know. i've been fairly diligent about using --reason (as well as deleting them as soon as i know i'm done with them) | 23:00 |
*** markvoelker has quit IRC | 23:00 | |
Shrews | jeblair: oh yeah. Thx | 23:00 |
clarkb | flaper87: if you are around I ran into a fail in the zaqar unittest suite on xenial. TypeError: 'NoneType' object has no attribute '__getitem__' not sure if known issue or not | 23:00 |
jeblair | Shrews: also, i'm digging your wip. :) | 23:01 |
Shrews | jeblair: it's more of an experiment to see how this design pans out. i think it might work | 23:02 |
jeblair | Shrews: it's very comprehensible | 23:03 |
Shrews | oh, you changed the topic already. danke | 23:03 |
*** shashank_hegde has quit IRC | 23:03 | |
*** csomerville has joined #openstack-infra | 23:03 | |
*** Sukhdev has joined #openstack-infra | 23:03 | |
*** shashank_hegde has joined #openstack-infra | 23:03 | |
jeblair | i made a little gertty dashboard of "status:open (branch:feature/zuulv3 OR topic:nodepool-zk)" | 23:04 |
Shrews | paradise by the little gertty dashboard light? | 23:05 |
openstackgerrit | Merged openstack-infra/zuul: Register connections when testing configuration https://review.openstack.org/300924 | 23:05 |
Shrews | sorry, i might be drinking now | 23:06 |
*** netsin has joined #openstack-infra | 23:06 | |
*** markvoelker has joined #openstack-infra | 23:06 | |
jeblair | Shrews: you should start work when i do, and i should start drinking when you do | 23:07 |
*** rdx_matt has joined #openstack-infra | 23:07 | |
Shrews | lol | 23:07 |
*** yaume has quit IRC | 23:07 | |
*** PollyZ has quit IRC | 23:07 | |
jeblair | notmorgan: can you look at 269731? i'm assuming it's harmless, and if we merge it in master, you'll just update it as needed -- but... i dunno? :) | 23:08 |
openstackgerrit | Merged openstack-infra/zuul: Don't take into account commit message for skip-if filter https://review.openstack.org/247383 | 23:08 |
openstackgerrit | Merged openstack-infra/zuul: Add additional zuul.conf information https://review.openstack.org/271868 | 23:10 |
*** rdx_matt has quit IRC | 23:12 | |
*** Sukhdev has quit IRC | 23:14 | |
Nakato | clarkb: re PBR WSGI. That's on my list to look at. | 23:14 |
clarkb | Nakato: yay, also not sure if you caught it but pifpaf (a tool jd__ wrote) does something similar and also fails | 23:15 |
clarkb | Nakato: so I wonder if this is a new exciting ubuntu xenial python2.7 behavior when reading pipes from subprocesses | 23:15 |
Nakato | clarkb: Oh, what patch is this that's failing? | 23:15 |
clarkb | Nakato: no patch, this is master | 23:16 |
Nakato | clarkb: The one I saw was related to the c extensions | 23:16 |
Nakato | Oh, interesting, hadn't seen that yet. | 23:16 |
*** karthikp_ has joined #openstack-infra | 23:16 | |
clarkb | Nakato: did you see my paste link of the full fail log? its in scrollback somewhere I Can dig it up if not | 23:16 |
openstackgerrit | Merged openstack-infra/zuul: highlight the error condition when reporting a config issue https://review.openstack.org/272262 | 23:17 |
openstackgerrit | Merged openstack-infra/zuul: Add quick-start guide https://review.openstack.org/271872 | 23:17 |
*** adrian_otto has joined #openstack-infra | 23:19 | |
*** shashank_hegde has quit IRC | 23:19 | |
openstackgerrit | Merged openstack-infra/zuul: Add extra test for bad url patterns https://review.openstack.org/260243 | 23:20 |
*** zeih has joined #openstack-infra | 23:20 | |
openstackgerrit | Merged openstack-infra/zuul: Clarify zuul-cloner message about missing branch https://review.openstack.org/259193 | 23:21 |
*** hemna is now known as hemnafk | 23:22 | |
*** xarses has quit IRC | 23:23 | |
openstackgerrit | Merged openstack-infra/zuul: Update documentation on statsd metrics generated. https://review.openstack.org/235521 | 23:24 |
*** zeih has quit IRC | 23:26 | |
openstackgerrit | Merged openstack-infra/zuul: Zuul references cleaner https://review.openstack.org/109276 | 23:27 |
openstackgerrit | Merged openstack-infra/zuul: Use high precedence when manually submitting jobs https://review.openstack.org/276406 | 23:27 |
*** dingyichen has joined #openstack-infra | 23:28 | |
openstackgerrit | Merged openstack-infra/zuul: Expose webapp listen_address and port https://review.openstack.org/261052 | 23:31 |
pabelanger | \o/ | 23:31 |
pabelanger | did that one a while ago | 23:31 |
Nakato | clarkb: I can't seem to find it in scrollback | 23:32 |
notmorgan | jeblair: looking | 23:32 |
notmorgan | jeblair: yeah that is harmless | 23:32 |
notmorgan | jeblair: should make pep8, etc happier for 2.5 | 23:33 |
fungi | Nakato: 21:24 utc | 23:33 |
clarkb | Nakato: http://paste.openstack.org/show/531023/ | 23:33 |
Nakato | Thanks! | 23:33 |
jeblair | notmorgan: thx; aprvd | 23:34 |
notmorgan | jeblair: i also left a comment on it re: py3 stuff. | 23:34 |
*** ccamacho has quit IRC | 23:34 | |
notmorgan | but it's good as is (i wouldn't change it) | 23:34 |
jeblair | jhesketh: do we want both https://review.openstack.org/229582 and https://review.openstack.org/262112 ? | 23:34 |
jeblair | notmorgan: agree with comment | 23:35 |
notmorgan | jeblair: once i get zuul v3 all py3 happy, are we going to do the same for nodepool, py3 only? | 23:35 |
notmorgan | or is nodepool going to want to sup[port py3/2 compat for a while/longtime/whatever | 23:35 |
notmorgan | no answer needed now | 23:35 |
notmorgan | just something to think about | 23:35 |
jeblair | notmorgan: i don't think it needs to support py2 -- but i'd prioritize that lower since it's not a zuulv3 blocker | 23:35 |
*** baoli_ has quit IRC | 23:35 | |
notmorgan | absolutely | 23:36 |
notmorgan | and it can totally just live on py2 for... well ever | 23:36 |
notmorgan | [aka until we get around to fixing it] | 23:36 |
jeblair | notmorgan: having said that -- greghaynes has a desire to use nodepool as an image building engine independent of zuul -- so we should see if py3-only is a hardship for that use case | 23:37 |
notmorgan | i doubt it will be, but yeah, we can play that by ear as we head closer to that | 23:37 |
*** Sukhdev has joined #openstack-infra | 23:38 | |
jeblair | (yeah, hopefully not since it's an anticipated future use case...) | 23:38 |
clarkb | notmorgan: jeblair one reason I could think of for that is rhel doesn't python3 iirc | 23:42 |
clarkb | so if you want an engine builder on centos/rhel you need python2? | 23:42 |
clarkb | looks like it is in epel and is a software collection though | 23:43 |
notmorgan | clarkb: epel should have it | 23:44 |
notmorgan | clarkb: yeah. | 23:44 |
*** hurgleburgler has quit IRC | 23:44 | |
jhesketh | jeblair: yes, they are subtly different | 23:44 |
jhesketh | jeblair: the latter is used in the swift work but could be useful where we want to give swift hte ability to push outside of the LOG_PATH but want to keep the LOG_PATH param otherwise | 23:45 |
*** tqtran has quit IRC | 23:46 | |
openstackgerrit | Marc Abramowitz proposed openstack-infra/jenkins-job-builder: Refactor delete_jobs https://review.openstack.org/88646 | 23:46 |
jeblair | jhesketh: and safe to merge both? | 23:47 |
jhesketh | jeblair: should be... one is behind the other.. but it was a long time since I looked | 23:47 |
*** amitgandhinz has joined #openstack-infra | 23:48 | |
*** Julien-zte has quit IRC | 23:48 | |
*** zz_dimtruck is now known as dimtruck | 23:51 | |
openstackgerrit | Patrick East proposed openstack-infra/devstack-gate: Add support for sub-node specific local.conf's https://review.openstack.org/341222 | 23:51 |
*** thorst has joined #openstack-infra | 23:51 | |
*** amitgandhinz has quit IRC | 23:53 | |
*** shashank_hegde has joined #openstack-infra | 23:53 | |
openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: WIP: ZooKeeper API https://review.openstack.org/340658 | 23:54 |
openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: Use random numbers for ZK test paths https://review.openstack.org/341223 | 23:54 |
*** EricGonczer_ has joined #openstack-infra | 23:56 | |
*** dizquierdo has quit IRC | 23:58 | |
*** sflanigan has joined #openstack-infra | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!