| oomichi | fungi: not google thing :) cncf is one fundation under the linux foundation | 00:00 |
|---|---|---|
| fungi | i thought it was where google was sending their open source efforts to make them seem less googly | 00:00 |
| oomichi | they have multiple projects including kubernetes, prometeus, etc | 00:00 |
| oomichi | fungi: yeah, that is kubernetes which is a main project under cncf | 00:01 |
| oomichi | and the project numer is still increasing under cncf by the other companies | 00:01 |
| oomichi | then they are interested in stackalytics which can show metrics by project easily | 00:02 |
| openstackgerrit | Paul Belanger proposed openstack-infra/devstack-gate master: Install stackviz from cached tarball https://review.openstack.org/457074 | 00:04 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Fix traceback when evaluating github changes https://review.openstack.org/470068 | 00:04 |
| fungi | oomichi: aha, i just read up on it again and apparently they've had a lot of additional projects "donated" by various companies since i originally heard about it | 00:05 |
| openstackgerrit | Paul Belanger proposed openstack-infra/devstack-gate master: Install stackviz from cached tarball https://review.openstack.org/457074 | 00:05 |
| *** pramodrj07 has quit IRC | 00:06 | |
| fungi | oomichi: i can see where they'd be interested in something like stackalytics since a majority of the cncf projects look like things we'd label single-vendor or non-diverse | 00:06 |
| fungi | no doubt they'd like to up the affiliation diversity of their contributor base on those | 00:07 |
| fungi | and encourage more cross-project collaboration | 00:07 |
| pabelanger | I mean, we can add cncf projects to stackalytics.o.o today if we wanted too | 00:08 |
| oomichi | fungi: yeah, that is a good point. The diversity is an important factor on open source, and they seem to want to take care of it | 00:08 |
| pabelanger | we do that some some other projects like ansible | 00:09 |
| pabelanger | http://stackalytics.openstack.org/?project_type=kubernetes-group&metric=commits | 00:09 |
| pabelanger | guess we track kubernetes right now | 00:09 |
| fungi | mrmartin has a spec and working code to fix the analytics cache persistence problems with stackalytics | 00:10 |
| *** asselin_ has joined #openstack-infra | 00:10 | |
| pabelanger | http://stackalytics.openstack.org/?project_type=docker-group&metric=commits is docker | 00:10 |
| pabelanger | fungi: ++ that would be helpful | 00:10 |
| oomichi | pabelanger: yeah, current stackalytics can show kubernetes one also. but I think it doesn't show it collectly to fit the latest situation | 00:11 |
| oomichi | now kubernetes has 41 repos under the project | 00:11 |
| fungi | i also am very close to being able to start work on tying stackalytics affiliation data into the openstack foundation member affiliation tracker so that people can update the affiliation in their profiles on www.openstack.org rather than adding configuration overrides in files in the stackalytics repo | 00:11 |
| oomichi | and its number still is increasing. It could be hard to follow it on openstack side | 00:12 |
| fungi | but all these major improvements will likely go nowhere if we can't get the current stackalytics maintainers on board with them | 00:12 |
| fungi | seems like time for another ml thread | 00:12 |
| *** xarses has quit IRC | 00:13 | |
| pabelanger | oomichi: I think we'd just need to update: http://git.openstack.org/cgit/openstack/stackalytics/tree/etc/default_data.json#n29680 | 00:13 |
| pabelanger | oomichi: but I'm not 100% | 00:13 |
| *** asselin__ has joined #openstack-infra | 00:13 | |
| fungi | look at the open stackalytics changes in gerrit... i keep getting cc'd on one adding a new "related technology" project there | 00:14 |
| *** asselin has quit IRC | 00:14 | |
| fungi | i get the impression even those aren't getting much review attention | 00:14 |
| *** yamamoto_ has joined #openstack-infra | 00:14 | |
| oomichi | pabelanger: yeah, you are right. That is just a config option, I am 100% sure :) | 00:16 |
| smcginnis | Can we just do a hostile takeover? :D | 00:16 |
| *** asselin_ has quit IRC | 00:16 | |
| fungi | smcginnis: not without an ml thread! ;) | 00:16 |
| oomichi | and that is a giant config file, https://review.openstack.org/#/c/469639 will help to make it small by separation | 00:19 |
| * oomichi finds way to ask reviewing :) | 00:20 | |
| *** stevemar__ has quit IRC | 00:22 | |
| *** stevemar__ has joined #openstack-infra | 00:23 | |
| *** rhallisey has joined #openstack-infra | 00:25 | |
| *** stevemar__ has quit IRC | 00:27 | |
| *** gyee has quit IRC | 00:30 | |
| *** dikonoor has joined #openstack-infra | 00:37 | |
| *** LindaWang has joined #openstack-infra | 00:39 | |
| *** jamielennox is now known as jamielennox|away | 00:44 | |
| *** salv-orlando has joined #openstack-infra | 00:45 | |
| *** cody-somerville has quit IRC | 00:45 | |
| *** ihrachys has joined #openstack-infra | 00:51 | |
| *** cuongnv has joined #openstack-infra | 00:52 | |
| *** salv-orlando has quit IRC | 00:54 | |
| *** ramishra has joined #openstack-infra | 01:03 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move global mount tracking into state https://review.openstack.org/469752 | 01:05 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 01:05 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Use global state to check for duplicate fs labels https://review.openstack.org/469751 | 01:05 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Remove 'state' argument from later cmd_* calls https://review.openstack.org/469750 | 01:05 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Add state to NodeBase class https://review.openstack.org/469749 | 01:05 |
| ianw | ^ i doubt i will ever get to write a better unit test than that in 464105 | 01:06 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 01:10 |
| *** thorst has joined #openstack-infra | 01:11 | |
| *** liujiong_lj has joined #openstack-infra | 01:19 | |
| *** liujiong_lj is now known as liujiong | 01:22 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-zuul master: Add ansible dependencies for zuulv3 https://review.openstack.org/470072 | 01:23 |
| *** bobh has quit IRC | 01:25 | |
| *** Apoorva_ has joined #openstack-infra | 01:27 | |
| *** Apoorva has quit IRC | 01:30 | |
| *** Apoorva_ has quit IRC | 01:31 | |
| *** tuanluong has joined #openstack-infra | 01:31 | |
| *** hongbin has joined #openstack-infra | 01:40 | |
| mikal | So... Does anyone have any advice on trying to nail down which unit test is attempting to do a sudo on a jenkins node? I haven't managed to repeat the problem on my home dev box. | 01:43 |
| mikal | The test of course just says "a sudo was attempted", not the test which was naughty | 01:43 |
| fungi | mikal: the check identifies it by looking for sudo attempts in the system auth log, so if you look there you can at least see what command it attempted to sudo | 01:46 |
| fungi | assuming there are a bounded number of tests being added/changed in the patch in question, hopefully that makes it easier to figure out which one is at fault | 01:47 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 01:47 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 01:47 |
| *** stevemar__ has joined #openstack-infra | 01:49 | |
| *** gomarivera has joined #openstack-infra | 01:51 | |
| *** rhallisey has quit IRC | 01:51 | |
| *** spzala has quit IRC | 01:52 | |
| *** dhinesh has quit IRC | 01:55 | |
| mikal | fungi: yeah, the problem is that the command is privsep-helper starting up, which all of the modified tests might do | 01:58 |
| mikal | I'm thinking maybe I patch the privsep code to throw an exception on startup | 01:59 |
| *** jamielennox|away is now known as jamielennox | 02:00 | |
| fungi | yeah, seems like you need something that can leave a better trail to follow... though in most cases where multiple tests are being added/modified and there's a sudo being blocked, you can bet it's a failing test (unless the test exercises something that tries to sudo and falls back on an alternative option) | 02:01 |
| *** thorst has quit IRC | 02:02 | |
| *** ihrachys has quit IRC | 02:04 | |
| *** huanxie has joined #openstack-infra | 02:05 | |
| *** spzala has joined #openstack-infra | 02:13 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move global mount tracking into state https://review.openstack.org/469752 | 02:14 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 02:14 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Use global state to check for duplicate fs labels https://review.openstack.org/469751 | 02:14 |
| *** liujiong has quit IRC | 02:14 | |
| *** liujiong_lj has joined #openstack-infra | 02:14 | |
| *** xinliang has quit IRC | 02:15 | |
| *** spzala has quit IRC | 02:18 | |
| *** links has joined #openstack-infra | 02:23 | |
| *** xinliang has joined #openstack-infra | 02:27 | |
| *** thorst has joined #openstack-infra | 02:29 | |
| *** gomarivera has quit IRC | 02:29 | |
| *** thorst has quit IRC | 02:30 | |
| *** stevemar__ has quit IRC | 02:33 | |
| *** faizy has quit IRC | 02:33 | |
| *** stevemar__ has joined #openstack-infra | 02:33 | |
| *** faizy has joined #openstack-infra | 02:34 | |
| *** stevemar__ has quit IRC | 02:38 | |
| *** mriedem has quit IRC | 02:40 | |
| *** thorst has joined #openstack-infra | 02:41 | |
| *** thorst has quit IRC | 02:41 | |
| *** spzala has joined #openstack-infra | 02:41 | |
| *** salv-orlando has joined #openstack-infra | 02:50 | |
| *** thorst has joined #openstack-infra | 02:51 | |
| *** thorst has quit IRC | 02:51 | |
| openstackgerrit | Merged openstack/diskimage-builder master: Adjust package mapping for SUSE family https://review.openstack.org/469533 | 02:56 |
| *** jascott1 has quit IRC | 02:59 | |
| *** jascott1 has joined #openstack-infra | 03:00 | |
| *** spzala has quit IRC | 03:01 | |
| *** salv-orlando has quit IRC | 03:02 | |
| *** jascott1 has quit IRC | 03:04 | |
| *** jamesmca_ has joined #openstack-infra | 03:05 | |
| mtreinish | andreaf: btw: https://review.openstack.org/#/c/469802 | 03:06 |
| mtreinish | andreaf: feel free to add onto that if you want to use your specific example | 03:06 |
| *** ramishra has quit IRC | 03:06 | |
| mtreinish | I didn't have any config files handy to use as a direct example | 03:07 |
| *** jamesmca_ has quit IRC | 03:09 | |
| *** bobh has joined #openstack-infra | 03:17 | |
| *** bobh has quit IRC | 03:18 | |
| *** gongysh has joined #openstack-infra | 03:22 | |
| *** spzala has joined #openstack-infra | 03:22 | |
| *** thorst has joined #openstack-infra | 03:22 | |
| *** cody-somerville has joined #openstack-infra | 03:23 | |
| *** spzala has quit IRC | 03:26 | |
| *** jamielennox is now known as jamielennox|away | 03:30 | |
| *** hongbin has quit IRC | 03:39 | |
| *** thorst has quit IRC | 03:40 | |
| *** armax has quit IRC | 03:41 | |
| *** armax has joined #openstack-infra | 03:45 | |
| *** psachin has joined #openstack-infra | 03:45 | |
| *** faizy_ has joined #openstack-infra | 03:47 | |
| *** nicolasbock has quit IRC | 03:48 | |
| *** faizy has quit IRC | 03:51 | |
| *** ramishra has joined #openstack-infra | 03:52 | |
| *** jascott1 has joined #openstack-infra | 03:54 | |
| *** sree has joined #openstack-infra | 03:55 | |
| *** vikrant has joined #openstack-infra | 03:56 | |
| *** armax has quit IRC | 04:00 | |
| *** ramishra has quit IRC | 04:05 | |
| *** faizy_ has quit IRC | 04:06 | |
| *** faizy has joined #openstack-infra | 04:07 | |
| *** gongysh has quit IRC | 04:07 | |
| *** ramishra has joined #openstack-infra | 04:15 | |
| *** aviau has quit IRC | 04:23 | |
| *** aviau has joined #openstack-infra | 04:24 | |
| *** salv-orlando has joined #openstack-infra | 04:25 | |
| *** stevemar__ has joined #openstack-infra | 04:34 | |
| *** salv-orlando has quit IRC | 04:37 | |
| *** thorst has joined #openstack-infra | 04:37 | |
| *** dhajare-brb has joined #openstack-infra | 04:37 | |
| *** stevemar__ has quit IRC | 04:39 | |
| *** jaosorior has joined #openstack-infra | 04:41 | |
| *** thorst has quit IRC | 04:42 | |
| *** knangia_ has quit IRC | 04:43 | |
| *** yamamoto_ has quit IRC | 04:46 | |
| *** kornicameister has joined #openstack-infra | 04:50 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 04:53 |
| *** ltomasbo|away is now known as ltomasbo | 04:53 | |
| *** ramishra has quit IRC | 04:57 | |
| *** ramishra has joined #openstack-infra | 05:01 | |
| *** dhinesh has joined #openstack-infra | 05:02 | |
| *** ramishra has quit IRC | 05:02 | |
| *** jascott1 has quit IRC | 05:06 | |
| *** ramishra has joined #openstack-infra | 05:07 | |
| *** claudiub has joined #openstack-infra | 05:09 | |
| openstackgerrit | Merged openstack-infra/project-config master: neutron: don't unstack services for functional and fullstack jobs https://review.openstack.org/470032 | 05:09 |
| *** jascott1 has joined #openstack-infra | 05:09 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Case sensitive label matching https://review.openstack.org/469946 | 05:11 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix tests for "Case sensitive label matching" https://review.openstack.org/469985 | 05:11 |
| *** ihrachys has joined #openstack-infra | 05:13 | |
| *** gouthamr has quit IRC | 05:15 | |
| *** rossella_s has joined #openstack-infra | 05:20 | |
| *** arahal_ has joined #openstack-infra | 05:21 | |
| *** ihrachys has quit IRC | 05:23 | |
| *** prateek has joined #openstack-infra | 05:23 | |
| *** rcernin has joined #openstack-infra | 05:26 | |
| *** eranrom_ has joined #openstack-infra | 05:27 | |
| andreaf | mtreinish: https://github.com/andreafrittoli/mqttwarn/tree/firehose_osxnotify | 05:27 |
| *** arahal_ has quit IRC | 05:27 | |
| *** markvoelker_ has joined #openstack-infra | 05:27 | |
| *** eranrom has quit IRC | 05:28 | |
| *** markvoelker has quit IRC | 05:28 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 05:28 |
| *** mmedvede has quit IRC | 05:28 | |
| *** mugsie has quit IRC | 05:29 | |
| *** mmedvede has joined #openstack-infra | 05:31 | |
| *** mugsie has joined #openstack-infra | 05:31 | |
| *** mugsie has quit IRC | 05:31 | |
| *** mugsie has joined #openstack-infra | 05:31 | |
| *** yamamoto_ has joined #openstack-infra | 05:32 | |
| *** salv-orlando has joined #openstack-infra | 05:33 | |
| *** fandi has joined #openstack-infra | 05:37 | |
| *** fandi has quit IRC | 05:37 | |
| *** fandi has joined #openstack-infra | 05:37 | |
| *** thorst has joined #openstack-infra | 05:38 | |
| *** gildub has quit IRC | 05:42 | |
| *** thorst has quit IRC | 05:43 | |
| openstackgerrit | Merged openstack-infra/project-config master: diskimage-builder/glean: Upgrade nodepool-src-* jobs to voting https://review.openstack.org/469538 | 05:49 |
| *** cshastri has joined #openstack-infra | 05:50 | |
| *** spzala has joined #openstack-infra | 05:50 | |
| *** csomerville has joined #openstack-infra | 05:51 | |
| *** pgadiya has joined #openstack-infra | 05:53 | |
| *** pbourke has joined #openstack-infra | 05:54 | |
| *** cody-somerville has quit IRC | 05:54 | |
| *** pbourke_ has quit IRC | 05:55 | |
| *** pgadiya has quit IRC | 05:55 | |
| *** spzala has quit IRC | 05:55 | |
| *** camunoz has quit IRC | 05:58 | |
| *** gildub has joined #openstack-infra | 05:59 | |
| *** aeng has quit IRC | 06:02 | |
| *** markus_z has joined #openstack-infra | 06:10 | |
| openstackgerrit | Merged openstack/diskimage-builder master: dhcp-all-interfaces.sh - Add support for InfiniBand interface DHCP https://review.openstack.org/465928 | 06:11 |
| *** spzala has joined #openstack-infra | 06:13 | |
| *** Swami has joined #openstack-infra | 06:15 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 06:15 |
| *** spzala has quit IRC | 06:18 | |
| *** esikache1 has joined #openstack-infra | 06:19 | |
| *** dikonoor has quit IRC | 06:25 | |
| *** dikonoor has joined #openstack-infra | 06:25 | |
| *** rossella_s has quit IRC | 06:26 | |
| *** lin_yang has quit IRC | 06:26 | |
| *** Swami has quit IRC | 06:32 | |
| *** salv-orlando has quit IRC | 06:35 | |
| strigazi | Hello, I have submitted a patch with a dependency in another patch. The dependency patch is merged but the second one still says Needs Verified | 06:35 |
| strigazi | Needs Verified: 462962 , dependency: 462887 | 06:35 |
| strigazi | What am I missing? | 06:36 |
| *** jascott1 has quit IRC | 06:37 | |
| *** CynicGalahad has joined #openstack-infra | 06:38 | |
| *** jascott1 has joined #openstack-infra | 06:38 | |
| *** dhinesh has quit IRC | 06:40 | |
| frickler | strigazi: I think you need to rebase, see the yellow dot next to the "Parent(s)" entry | 06:41 |
| strigazi | frickler What the yellow dot means? | 06:42 |
| strigazi | The parent is merged | 06:42 |
| strigazi | I hoverd, not current, possible rebase | 06:42 |
| frickler | strigazi: 462962 is based on an old PS of 462887 | 06:43 |
| *** jascott1 has quit IRC | 06:43 | |
| strigazi | frickler yeap got it | 06:43 |
| *** dhajare-brb has quit IRC | 06:44 | |
| *** ralonsoh has joined #openstack-infra | 06:45 | |
| *** matrohon has joined #openstack-infra | 06:48 | |
| *** spzala has joined #openstack-infra | 06:49 | |
| *** flepied has joined #openstack-infra | 06:50 | |
| *** florianf has joined #openstack-infra | 06:50 | |
| *** spzala has quit IRC | 06:54 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into plugin_base object https://review.openstack.org/464105 | 06:57 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move rollback into NodeBase object https://review.openstack.org/464105 | 06:58 |
| *** rattboi has left #openstack-infra | 07:00 | |
| *** fandi has quit IRC | 07:01 | |
| *** tesseract has joined #openstack-infra | 07:04 | |
| *** dmellado_ has joined #openstack-infra | 07:05 | |
| *** dmellado_ has quit IRC | 07:05 | |
| *** jamesmca_ has joined #openstack-infra | 07:06 | |
| *** jamesmca_ has quit IRC | 07:10 | |
| *** rwsu has quit IRC | 07:28 | |
| *** dikonoor has quit IRC | 07:28 | |
| *** amoralej|off is now known as amoralej | 07:29 | |
| *** dhajare-brb has joined #openstack-infra | 07:30 | |
| *** jpena|off is now known as jpena | 07:34 | |
| *** eranrom_ is now known as eranrom | 07:35 | |
| *** electrofelix has joined #openstack-infra | 07:35 | |
| *** thorst has joined #openstack-infra | 07:39 | |
| *** thorst has quit IRC | 07:44 | |
| *** jpich has joined #openstack-infra | 07:45 | |
| *** hashar has joined #openstack-infra | 07:49 | |
| *** jascott1 has joined #openstack-infra | 07:55 | |
| *** dtantsur|afk has quit IRC | 07:59 | |
| *** jascott1 has quit IRC | 07:59 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-infra | 08:01 | |
| *** dtantsur has joined #openstack-infra | 08:03 | |
| *** Hal has quit IRC | 08:05 | |
| *** prashkre has joined #openstack-infra | 08:06 | |
| *** faizy has quit IRC | 08:08 | |
| *** sree has quit IRC | 08:08 | |
| *** sree has joined #openstack-infra | 08:09 | |
| *** faizy has joined #openstack-infra | 08:10 | |
| *** faizy has quit IRC | 08:12 | |
| *** faizy has joined #openstack-infra | 08:13 | |
| *** sree has quit IRC | 08:13 | |
| *** neiljerram has joined #openstack-infra | 08:15 | |
| openstackgerrit | Alejandro Andreu proposed openstack-infra/project-config master: Temporarily disable Beaker jobs for puppet-midonet https://review.openstack.org/470192 | 08:16 |
| *** sree has joined #openstack-infra | 08:17 | |
| *** lucas-afk is now known as lucasagomes | 08:18 | |
| *** prashkre has quit IRC | 08:18 | |
| *** Hal has joined #openstack-infra | 08:20 | |
| *** jamielennox|away is now known as jamielennox | 08:21 | |
| *** mhickey has joined #openstack-infra | 08:29 | |
| *** dikonoor has joined #openstack-infra | 08:29 | |
| openstackgerrit | Emilien Macchi proposed openstack-infra/project-config master: tripleo: add rule for gate-tripleo-ci-centos-7-containers-multinode-upgrades https://review.openstack.org/470200 | 08:33 |
| EmilienM | jistr: ^ here we go ^ I think being more specific like this can't hurt | 08:33 |
| jistr | EmilienM: right it shouldn't hurt, but i'm still puzzled how come this doesn't do the job right? https://github.com/openstack-infra/project-config/blob/85b22379ffa47bc7cc9c10efccad85066a20bfcc/zuul/layout.yaml#L1714-L1716 | 08:35 |
| jistr | EmilienM: can you post a link on which patch did you see the upgrade job run on Ocata? | 08:35 |
| *** derekh has joined #openstack-infra | 08:35 | |
| EmilienM | jistr: https://review.openstack.org/#/c/470164 | 08:36 |
| *** stevemar__ has joined #openstack-infra | 08:37 | |
| AJaeger | jistr: the layout log file explains which branch conditions are applied to a job, if in doubt double check what's there and whether that's correct. | 08:38 |
| *** dingyichen has quit IRC | 08:39 | |
| AJaeger | EmilienM: suggestion for simplication added | 08:40 |
| jistr | AJaeger: where can i find the layout log file? | 08:40 |
| *** thorst has joined #openstack-infra | 08:40 | |
| EmilienM | AJaeger: danke | 08:41 |
| * jistr looking at http://logs.openstack.org/64/470164/1/check/gate-tripleo-ci-centos-7-containers-multinode-upgrades-nv/7efde39/ | 08:41 | |
| EmilienM | AJaeger: that's weird, I didn't see this rule | 08:41 |
| EmilienM | AJaeger: and it does exactly what we want | 08:41 |
| *** ggnel_t has joined #openstack-infra | 08:42 | |
| *** stevemar__ has quit IRC | 08:42 | |
| openstackgerrit | YAMAMOTO Takashi proposed openstack-infra/project-config master: Retire gate-neutron-dsvm-tempest-vpnaas-ubuntu-trusty-nv https://review.openstack.org/470206 | 08:42 |
| EmilienM | AJaeger: gate-tripleo-ci-centos-7-containers-multinode-upgrades-nv shouldn't run on stable/ocata | 08:42 |
| AJaeger | EmilienM: let's see why not... | 08:42 |
| EmilienM | AJaeger: ah it doesn't have container keyword in the name | 08:43 |
| AJaeger | EmilienM: currently "^(?!stable/newton)." is used as branch condition - so I agree, a change is needed ;) | 08:44 |
| *** alexchadin has joined #openstack-infra | 08:46 | |
| AJaeger | EmilienM: there's condition that is too early - let's remove that one and I'm happy ;) | 08:47 |
| *** b00tcat_ has joined #openstack-infra | 08:48 | |
| AJaeger | EmilienM: could you review https://review.openstack.org/470192 , please? | 08:48 |
| EmilienM | sure | 08:48 |
| *** jamielennox is now known as jamielennox|away | 08:49 | |
| b00tcat_ | thanks EmilienM and AJaeger, I was just going to ask for reviews :-) | 08:49 |
| AJaeger | b00tcat_: let's propose a revert directly as well | 08:50 |
| EmilienM | AJaeger: I'm confused why L1715 doesn't work | 08:51 |
| EmilienM | AJaeger: I would rather keep this rule, in case. Because we have other jobs with "container" string in the jobname | 08:52 |
| openstackgerrit | Alejandro Andreu proposed openstack-infra/project-config master: WIP: Re-enable beaker tests for puppet-midonet https://review.openstack.org/470210 | 08:53 |
| openstackgerrit | Spyros Trigazis (strigazi) proposed openstack-infra/project-config master: [magnum] Use install deps script from DC/OS dir https://review.openstack.org/469366 | 08:54 |
| *** liujiong_lj has quit IRC | 08:57 | |
| *** liujiong has joined #openstack-infra | 08:59 | |
| *** thorst has quit IRC | 08:59 | |
| AJaeger | EmilienM: that regex was introduced with the job - so, let's not keep it if it's not used. | 09:02 |
| *** dizquierdo has joined #openstack-infra | 09:02 | |
| AJaeger | I think line 1859 is the problem | 09:02 |
| AJaeger | "name: ^gate-tripleo-ci-centos-7.*-upgrades.*$" | 09:02 |
| EmilienM | ok | 09:02 |
| AJaeger | we could move line 1715 below 1859 | 09:02 |
| EmilienM | ok | 09:04 |
| *** grosjean has left #openstack-infra | 09:05 | |
| *** dikonoor has quit IRC | 09:05 | |
| *** sambetts|afk is now known as sambetts | 09:06 | |
| *** jamesmca_ has joined #openstack-infra | 09:06 | |
| openstackgerrit | Emilien Macchi proposed openstack-infra/project-config master: tripleo: add rule for gate-tripleo-ci-centos-7-containers-multinode-upgrades https://review.openstack.org/470200 | 09:07 |
| EmilienM | AJaeger: like this is enough? ^ | 09:07 |
| *** neiljerram has quit IRC | 09:08 | |
| AJaeger | EmilienM: just moving around should be enough | 09:09 |
| openstackgerrit | Jiri Stransky proposed openstack-infra/tripleo-ci master: Add new container jobs to cistatus.html https://review.openstack.org/470213 | 09:09 |
| EmilienM | AJaeger: ok so I remove the new line that I created, right? | 09:10 |
| EmilienM | oh you commented | 09:10 |
| EmilienM | oki | 09:10 |
| EmilienM | AJaeger: thanks for your help today | 09:10 |
| openstackgerrit | Emilien Macchi proposed openstack-infra/project-config master: tripleo: add rule for gate-tripleo-ci-centos-7-containers-multinode-upgrades https://review.openstack.org/470200 | 09:10 |
| openstackgerrit | Merged openstack-infra/project-config master: [magnum] Use install deps script from DC/OS dir https://review.openstack.org/469366 | 09:10 |
| *** neiljerram has joined #openstack-infra | 09:10 | |
| *** jamesmca_ has quit IRC | 09:11 | |
| AJaeger | EmilienM: LGTM, let's wait for test results to confirm ;) | 09:14 |
| EmilienM | yeah cool | 09:14 |
| EmilienM | no pressure | 09:14 |
| AJaeger | EmilienM: checked logs, looks good now ;) | 09:19 |
| EmilienM | yeah ! | 09:19 |
| *** ccamacho is now known as ccamacho|brb | 09:19 | |
| *** faizy has quit IRC | 09:21 | |
| *** s-shiono has quit IRC | 09:21 | |
| *** e0ne has joined #openstack-infra | 09:23 | |
| *** sree has quit IRC | 09:30 | |
| *** sree has joined #openstack-infra | 09:31 | |
| *** sree has quit IRC | 09:35 | |
| *** dikonoor has joined #openstack-infra | 09:36 | |
| *** persia has quit IRC | 09:38 | |
| *** persia has joined #openstack-infra | 09:39 | |
| *** dizquierdo has quit IRC | 09:40 | |
| *** wolverineav has joined #openstack-infra | 09:42 | |
| openstackgerrit | Pavel Goglov proposed openstack-infra/jenkins-job-builder master: Add pipeline/workflow multibranch module https://review.openstack.org/403940 | 09:49 |
| *** ccamacho|brb is now known as ccamacho|off | 09:53 | |
| *** e0ne has quit IRC | 09:55 | |
| *** stakeda has quit IRC | 09:55 | |
| *** dizquierdo has joined #openstack-infra | 09:55 | |
| yolanda | AJaeger, around? i'm hitting a problem with a job, tox seems to be cutting the path of the directory to install: http://logs.openstack.org/11/470211/1/experimental/gate-tripleo-buildimage-overcloud-security-hardened-full-centos-7-nv/9a725d4/console.html#_2017-06-02_09_22_32_175369 | 09:55 |
| *** MatthewLi has quit IRC | 09:56 | |
| yolanda | how could i solve it? the job gets created with the name of my image, that is very long... but i wouldn't like to change the image name for tox, can you think on a better way? | 09:56 |
| *** thorst has joined #openstack-infra | 09:56 | |
| AJaeger | yolanda: let me check... | 09:57 |
| AJaeger | yolanda: linux kernel has a limit of 127 chars for command invocation, and tox often hits that with overlong job names . Shortening the name is the only option you have | 09:58 |
| yolanda | AJaeger, and is there anyway i can just shorten the name of the job, but not for the image name itself? | 09:58 |
| AJaeger | yolanda: you could have two parameter instead of one - so not hardcode the full name into the job | 09:59 |
| AJaeger | yolanda: FYI, I'll be offline the next two weeks on vacation without internet... | 10:00 |
| yolanda | ouch, going to miss you | 10:00 |
| *** aarefiev_afk is now known as aarefiev | 10:00 | |
| *** thorst has quit IRC | 10:01 | |
| yolanda | AJaeger, so the template is defined as name: 'gate-tripleo-buildimage-{image-name}-{node}{suffix}' | 10:01 |
| yolanda | what do you mean with 2 parameters? | 10:01 |
| AJaeger | let me check... | 10:01 |
| *** jaosorior is now known as jaosorior_lunch | 10:01 | |
| AJaeger | I mean: | 10:02 |
| AJaeger | gate-tripleo-buildimage-{short-name}-{node}{suffix}' | 10:02 |
| AJaeger | and then in projects.yaml you pass in both short-name and image-name (long name) | 10:02 |
| yolanda | mm, the problem i can see, is that i'd need to do that for all the jobs using that template | 10:03 |
| *** nicolasbock has joined #openstack-infra | 10:03 | |
| yolanda | AJaeger, also what worries me is if 127 char limit is a hard limitation, i could hit same issues with other jobs for example packaging... | 10:04 |
| yolanda | if that's not a limitation from our side, but from the kernel | 10:04 |
| *** sree has joined #openstack-infra | 10:04 | |
| AJaeger | yolanda: we hit this every few months - in a tox based job | 10:05 |
| yolanda | so this time i was the winner :) | 10:05 |
| *** qba73 has joined #openstack-infra | 10:05 | |
| AJaeger | yolanda: for all the other jobs, you could use "short-name" and "image-name" as the same value ;) | 10:05 |
| yolanda | i think it will be better to shorten it, i don't want to hit the same on rdo jobs for example | 10:05 |
| *** jamesmca_ has joined #openstack-infra | 10:05 | |
| AJaeger | but yeah, it's a change... | 10:05 |
| *** e0ne has joined #openstack-infra | 10:06 | |
| *** ericyoung has joined #openstack-infra | 10:07 | |
| b00tcat_ | hi yolanda, could you please have a quick look at https://review.openstack.org/#/c/470192/ ? | 10:07 |
| *** lunarlamp has quit IRC | 10:08 | |
| *** sree has quit IRC | 10:08 | |
| *** ykarel has joined #openstack-infra | 10:08 | |
| *** jamesmca_ has quit IRC | 10:10 | |
| yolanda | ok done | 10:11 |
| *** mariusv has joined #openstack-infra | 10:11 | |
| b00tcat_ | thanks! | 10:12 |
| *** wfoster has quit IRC | 10:12 | |
| *** rook has quit IRC | 10:12 | |
| *** lucasagomes has quit IRC | 10:13 | |
| *** tumbarka has quit IRC | 10:13 | |
| *** tushar has joined #openstack-infra | 10:13 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/project-config master: Shorten the image name for the security hardened images, because we were hitting the hard limit of 127 chars for directory names in jobs. https://review.openstack.org/470236 | 10:14 |
| openstackgerrit | yolanda.robla proposed openstack-infra/project-config master: Shorten the image name for the security hardened images https://review.openstack.org/470236 | 10:15 |
| openstackgerrit | Merged openstack-infra/project-config master: Temporarily disable Beaker jobs for puppet-midonet https://review.openstack.org/470192 | 10:18 |
| *** liujiong has quit IRC | 10:19 | |
| *** salv-orlando has joined #openstack-infra | 10:20 | |
| *** yamamoto_ has quit IRC | 10:22 | |
| *** rook has joined #openstack-infra | 10:27 | |
| *** rook is now known as Guest92604 | 10:27 | |
| *** lucasagomes has joined #openstack-infra | 10:29 | |
| *** lucasagomes has quit IRC | 10:30 | |
| *** jkilpatr has quit IRC | 10:30 | |
| *** lucasagomes has joined #openstack-infra | 10:31 | |
| *** wfoster has joined #openstack-infra | 10:31 | |
| *** cuongnv has quit IRC | 10:32 | |
| *** tuanluong has quit IRC | 10:35 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Make finger port configurable https://review.openstack.org/470243 | 10:38 |
| *** alexchadin has quit IRC | 10:41 | |
| b00tcat_ | yolanda: how long does it take for the new config to be actually in-place? somehow after commenting out the beaker tests line, it's still being triggered on the patch I submitted | 10:42 |
| *** ltomasbo is now known as ltomasbo|away | 10:42 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Don't copy the __pycache__ folder https://review.openstack.org/470245 | 10:42 |
| *** mariusv has quit IRC | 10:44 | |
| *** rossella_s has joined #openstack-infra | 10:45 | |
| *** mat128 has joined #openstack-infra | 10:45 | |
| *** lucasagomes has quit IRC | 10:46 | |
| *** mariusv has joined #openstack-infra | 10:46 | |
| b00tcat_ | ah no, somehow it disappeared from the current run in zuul, thanks anyway! | 10:46 |
| *** lucasagomes has joined #openstack-infra | 10:47 | |
| *** jkilpatr has joined #openstack-infra | 10:53 | |
| *** thorst has joined #openstack-infra | 10:57 | |
| *** gildub has quit IRC | 10:57 | |
| *** LindaWang has quit IRC | 10:59 | |
| *** thorst has quit IRC | 11:02 | |
| *** e0ne has quit IRC | 11:02 | |
| *** dizquierdo has quit IRC | 11:03 | |
| *** dave-mccowan has joined #openstack-infra | 11:03 | |
| *** dizquierdo has joined #openstack-infra | 11:04 | |
| *** matrohon has quit IRC | 11:04 | |
| *** mhickey has quit IRC | 11:06 | |
| *** lucasagomes is now known as lucas-hungry | 11:06 | |
| *** neiljerram has quit IRC | 11:12 | |
| *** dikonoo has joined #openstack-infra | 11:13 | |
| *** dikonoor has quit IRC | 11:16 | |
| *** dklyle has quit IRC | 11:25 | |
| *** esikache1 has quit IRC | 11:26 | |
| *** esikachev has joined #openstack-infra | 11:26 | |
| *** fandi has joined #openstack-infra | 11:28 | |
| *** bobh has joined #openstack-infra | 11:29 | |
| *** salv-orlando has quit IRC | 11:29 | |
| *** bobh has quit IRC | 11:30 | |
| *** david-lyle has joined #openstack-infra | 11:31 | |
| *** gcb has quit IRC | 11:34 | |
| *** matrohon has joined #openstack-infra | 11:36 | |
| *** rossella_s has quit IRC | 11:38 | |
| *** gcb has joined #openstack-infra | 11:38 | |
| *** thorst has joined #openstack-infra | 11:41 | |
| *** rhallisey has joined #openstack-infra | 11:42 | |
| *** gcb has quit IRC | 11:43 | |
| openstackgerrit | Pavel Goglov proposed openstack-infra/jenkins-job-builder master: Add pipeline/workflow multibranch module https://review.openstack.org/403940 | 11:44 |
| *** thorst has quit IRC | 11:45 | |
| *** nicolasbock has quit IRC | 11:46 | |
| *** jpena is now known as jpena|lunch | 11:47 | |
| openstackgerrit | Dirk Mueller proposed openstack-infra/devstack-gate master: Install systemd-journal-remote earlier https://review.openstack.org/465909 | 11:47 |
| *** yamamoto has joined #openstack-infra | 11:47 | |
| openstackgerrit | Merged openstack-infra/project-config master: tripleo: add rule for gate-tripleo-ci-centos-7-containers-multinode-upgrades https://review.openstack.org/470200 | 11:49 |
| openstackgerrit | Dirk Mueller proposed openstack-infra/devstack-gate master: openSUSE enablement for devstack-gate https://review.openstack.org/468685 | 11:49 |
| openstackgerrit | Dirk Mueller proposed openstack-infra/devstack-gate master: openSUSE enablement for devstack-gate https://review.openstack.org/468685 | 11:51 |
| *** gcb has joined #openstack-infra | 11:52 | |
| openstackgerrit | Merged openstack-infra/project-config master: Running volume multibackend slow scenario tests in scenario job https://review.openstack.org/469334 | 11:52 |
| openstackgerrit | Merged openstack-infra/openstack-zuul-jobs master: Collect testr / testrepository results https://review.openstack.org/469947 | 11:52 |
| *** stevemar__ has joined #openstack-infra | 11:55 | |
| *** nicolasbock has joined #openstack-infra | 11:56 | |
| *** ldnunes has joined #openstack-infra | 11:56 | |
| openstackgerrit | Merged openstack-infra/project-config master: Add new hosted project: openstack/ptgbot https://review.openstack.org/469928 | 11:59 |
| *** rfolco has joined #openstack-infra | 12:01 | |
| *** thorst has joined #openstack-infra | 12:01 | |
| openstackgerrit | Merged openstack-infra/project-config master: Add new project(fuxi-golang) to Openstack https://review.openstack.org/468635 | 12:02 |
| openstackgerrit | Merged openstack-infra/project-config master: Add Multinode gate for OpenStack-Helm project https://review.openstack.org/466287 | 12:04 |
| *** sree has joined #openstack-infra | 12:05 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Make finger port configurable https://review.openstack.org/470243 | 12:09 |
| *** sree has quit IRC | 12:09 | |
| openstackgerrit | Pavel Goglov proposed openstack-infra/jenkins-job-builder master: Add pipeline/workflow multibranch module https://review.openstack.org/403940 | 12:11 |
| *** alegacy has left #openstack-infra | 12:13 | |
| *** neith has joined #openstack-infra | 12:17 | |
| neith | hello | 12:17 |
| neith | I was sent here from regarding an issue to sign the OpenStack Individual Contributor License Agreement | 12:18 |
| neith | It always fails | 12:18 |
| neith | with an error: Code Review - Error | 12:19 |
| neith | Server Error | 12:19 |
| neith | Cannot store contact information | 12:19 |
| neith | anyone can help? | 12:19 |
| openstackgerrit | Merged openstack-infra/shade master: Remove designateclient from commands related to recordsets https://review.openstack.org/469997 | 12:21 |
| *** lucas-hungry is now known as lucasagomes | 12:23 | |
| *** baoli has joined #openstack-infra | 12:28 | |
| *** baoli has quit IRC | 12:31 | |
| *** baoli has joined #openstack-infra | 12:31 | |
| *** nicolasbock has quit IRC | 12:33 | |
| *** florianf has quit IRC | 12:34 | |
| *** dizquierdo has quit IRC | 12:36 | |
| openstackgerrit | Vitaly Gridnev proposed openstack-infra/project-config master: [sahara] grenade jobs changes https://review.openstack.org/470271 | 12:39 |
| *** dpawlik has quit IRC | 12:39 | |
| *** trown|outtypewww is now known as trown | 12:40 | |
| *** ykarel has quit IRC | 12:40 | |
| *** rlandy has joined #openstack-infra | 12:40 | |
| *** spzala has joined #openstack-infra | 12:41 | |
| *** annegentle has joined #openstack-infra | 12:41 | |
| *** mriedem has joined #openstack-infra | 12:43 | |
| *** yamamoto has quit IRC | 12:43 | |
| *** jpena|lunch is now known as jpena | 12:47 | |
| *** stevemar__ has quit IRC | 12:47 | |
| *** dprince has joined #openstack-infra | 12:49 | |
| *** florianf has joined #openstack-infra | 12:49 | |
| *** yamamoto has joined #openstack-infra | 12:50 | |
| *** stevemar__ has joined #openstack-infra | 12:51 | |
| *** ramishra has quit IRC | 12:54 | |
| *** tdurakov has quit IRC | 12:54 | |
| *** annegentle has quit IRC | 12:56 | |
| *** ramishra has joined #openstack-infra | 12:56 | |
| *** stevemar__ has quit IRC | 12:56 | |
| *** Guest92604 is now known as rook | 12:59 | |
| *** rook is now known as Guest53691 | 12:59 | |
| *** yamamoto has quit IRC | 13:00 | |
| *** gordc has joined #openstack-infra | 13:01 | |
| *** jascott1 has joined #openstack-infra | 13:01 | |
| *** e0ne has joined #openstack-infra | 13:02 | |
| *** jaosorior_lunch is now known as jaosorior | 13:03 | |
| *** gcb has quit IRC | 13:04 | |
| *** MatthewLi has joined #openstack-infra | 13:04 | |
| *** vikrant has quit IRC | 13:05 | |
| *** ramishra has quit IRC | 13:05 | |
| *** jascott1 has quit IRC | 13:06 | |
| *** salv-orlando has joined #openstack-infra | 13:07 | |
| *** ykarel has joined #openstack-infra | 13:08 | |
| *** Hal has quit IRC | 13:09 | |
| openstackgerrit | Merged openstack-infra/shade master: Add logging of non-standard error message documents https://review.openstack.org/468092 | 13:09 |
| openstackgerrit | Merged openstack-infra/shade master: Set some logger names explicitly https://review.openstack.org/468093 | 13:09 |
| *** Hal has joined #openstack-infra | 13:09 | |
| *** askb has quit IRC | 13:09 | |
| *** kgiusti has joined #openstack-infra | 13:10 | |
| *** mdrabe has joined #openstack-infra | 13:10 | |
| *** esikachev has quit IRC | 13:11 | |
| *** nicolasbock has joined #openstack-infra | 13:13 | |
| openstackgerrit | Merged openstack-infra/shade master: Add novaclient interactions to http_debug https://review.openstack.org/468094 | 13:14 |
| AJaeger | b00tcat_: until the next puppet run - so, in general should be active after an hour | 13:14 |
| AJaeger | neith: are you a foundation member? | 13:15 |
| *** esikachev has joined #openstack-infra | 13:15 | |
| AJaeger | neith: This needs your gerrit preferred e-mail address to match a primary | 13:16 |
| AJaeger | e-mail address for a foundation individual member account. | 13:16 |
| AJaeger | If you already followed the instructions at | 13:16 |
| AJaeger | http://docs.openstack.org/infra/manual/developers.html#account-setup | 13:16 |
| AJaeger | and still get that, see https://ask.openstack.org/question/56720 for | 13:16 |
| AJaeger | additional troubleshooting tips. | 13:16 |
| neith | AJaeger: it matches | 13:16 |
| neith | its the same | 13:16 |
| neith | AJaeger: community member | 13:17 |
| AJaeger | neith: that's your problem - you need foundatoin member, see the two links I mentioned for details | 13:18 |
| *** esikachev has quit IRC | 13:20 | |
| neith | AJaeger: ohhh | 13:20 |
| neith | AJaeger: alright | 13:20 |
| neith | thks | 13:20 |
| neith | it works now | 13:20 |
| *** jklare_ has joined #openstack-infra | 13:22 | |
| *** jklare has quit IRC | 13:22 | |
| *** jklare_ is now known as jklare | 13:22 | |
| *** trown is now known as trown|brb | 13:23 | |
| *** ihrachys has joined #openstack-infra | 13:24 | |
| *** Dinesh_Bhor has quit IRC | 13:24 | |
| AJaeger | neith: glad to hear | 13:25 |
| *** psachin has quit IRC | 13:26 | |
| *** pradk has quit IRC | 13:27 | |
| *** Goneri has joined #openstack-infra | 13:27 | |
| *** prateek has quit IRC | 13:27 | |
| *** salv-orl_ has joined #openstack-infra | 13:27 | |
| *** trown|brb is now known as trown | 13:29 | |
| *** salv-orlando has quit IRC | 13:30 | |
| openstackgerrit | Chandan Kumar proposed openstack-infra/project-config master: Switch to bash script for getting tempest plugin list https://review.openstack.org/470286 | 13:30 |
| openstackgerrit | Vitaly Gridnev proposed openstack-infra/project-config master: [sahara] grenade jobs changes https://review.openstack.org/470271 | 13:32 |
| *** esikachev has joined #openstack-infra | 13:33 | |
| *** fandi has quit IRC | 13:34 | |
| *** links has quit IRC | 13:35 | |
| *** esikachev has quit IRC | 13:37 | |
| openstackgerrit | Evgeny Antyshev proposed openstack-infra/puppet-pip master: Don't install pip3 if no python3 installed https://review.openstack.org/470289 | 13:38 |
| *** amoralej is now known as amoralej|lunch | 13:39 | |
| *** esikachev has joined #openstack-infra | 13:39 | |
| *** arahal_ has joined #openstack-infra | 13:39 | |
| *** yland has joined #openstack-infra | 13:40 | |
| *** felipemonteiro has joined #openstack-infra | 13:46 | |
| *** pradk has joined #openstack-infra | 13:47 | |
| *** yland has quit IRC | 13:48 | |
| *** yland has joined #openstack-infra | 13:48 | |
| *** yland is now known as Guest75163 | 13:49 | |
| *** e0ne has quit IRC | 13:49 | |
| *** CynicGalahad has quit IRC | 13:49 | |
| *** gomarivera has joined #openstack-infra | 13:50 | |
| *** stevemar__ has joined #openstack-infra | 13:51 | |
| *** esikachev has quit IRC | 13:51 | |
| *** gouthamr has joined #openstack-infra | 13:51 | |
| *** felipemonteiro__ has joined #openstack-infra | 13:51 | |
| *** xarses has joined #openstack-infra | 13:51 | |
| *** xarses has quit IRC | 13:52 | |
| *** xarses has joined #openstack-infra | 13:52 | |
| chandankumar | AJaeger: hello | 13:54 |
| *** felipemonteiro has quit IRC | 13:55 | |
| chandankumar | AJaeger: regarding this review https://review.openstack.org/#/c/466630/ , will i also move the zuul cloner and tempest part with in tempest? | 13:55 |
| *** esikachev has joined #openstack-infra | 13:55 | |
| chandankumar | AJaeger: and in ci i will only run tox -e sanity that will execute stuff ? | 13:55 |
| lucasagomes | hi all, have anyone seem this with reno 2.3.1 ? http://logs.openstack.org/03/461103/11/check/gate-networking-ovn-releasenotes/8de0c7d/console.html#_2017-06-02_12_13_31_803258 | 13:59 |
| lucasagomes | downgrading it to 2.3.0 the job (locally) passes | 13:59 |
| *** yamamoto has joined #openstack-infra | 14:01 | |
| *** hrubi has quit IRC | 14:03 | |
| pabelanger | mordred: jeblair: clarkb: fungi: puppet-zuul patch to add ssl / ffi headers needed to compile ansible: https://review.openstack.org/#/c/470072/ | 14:03 |
| *** hrubi has joined #openstack-infra | 14:04 | |
| *** eharney has joined #openstack-infra | 14:05 | |
| *** yamamoto has quit IRC | 14:06 | |
| *** yamamoto has joined #openstack-infra | 14:07 | |
| *** rbrndt has joined #openstack-infra | 14:07 | |
| openstackgerrit | Evgeny Antyshev proposed openstack-infra/puppet-pip master: Don't install pip3 if no python3 installed https://review.openstack.org/470289 | 14:08 |
| pabelanger | lucasagomes: might want to confirm with dhellmann | 14:08 |
| lucasagomes | pabelanger, yeah apparently 2.3.1 last commit was from him... I'm just trying to figure out why it's failing due that last change | 14:09 |
| lucasagomes | dhellmann, if you have some time, any clue on that reno failure above ^ ? | 14:09 |
| *** jamesmca_ has joined #openstack-infra | 14:10 | |
| *** MatthewLi has quit IRC | 14:12 | |
| *** dikonoo has quit IRC | 14:14 | |
| *** baoli_ has joined #openstack-infra | 14:15 | |
| *** e0ne has joined #openstack-infra | 14:18 | |
| *** baoli has quit IRC | 14:18 | |
| fungi | pabelanger: are libffi and libssl being linked directly into ansible or into dependencies of ansible? if it's the latter, we're already installing most of the other python dependencies as system packages so could presumably do the same for consistency? | 14:19 |
| *** stevemar__ has quit IRC | 14:19 | |
| *** dizquierdo has joined #openstack-infra | 14:19 | |
| fungi | i.e. python-cffi and python-openssl | 14:20 |
| *** amoralej|lunch is now known as amoralej | 14:20 | |
| *** markus_z has quit IRC | 14:21 | |
| lucasagomes | pabelanger, dhellmann this https://review.openstack.org/#/c/469966/ does seem to fix it (tested locally and it works) | 14:21 |
| pabelanger | fungi: I belive it is cryptography that is needed them, which is pulled in by ansible | 14:21 |
| openstackgerrit | Derek Higgins proposed openstack-infra/project-config master: WIP: Add job to test ironic boot from volume https://review.openstack.org/470307 | 14:22 |
| pabelanger | Ya, cryptography needs them | 14:22 |
| fungi | pabelanger: right, just wondering why the mix of preinstalling some python deps as system packages but compiling others from sdist | 14:22 |
| *** aarefiev is now known as aarefiev_afk | 14:22 | |
| pabelanger | fungi: I am not sure to be honest | 14:23 |
| pabelanger | it is not very clean currently | 14:24 |
| *** stevemar__ has joined #openstack-infra | 14:25 | |
| *** csomerville has quit IRC | 14:25 | |
| pabelanger | if we are doing pip install, I would imaging system python packages would not be used. Since we'd have no control over pip overwriting them | 14:25 |
| fungi | current master branch of pyca/cryptography claims to need cffi>=1.7 which isn't satisfyable on ubuntu until yakkety, abd python-cryptography itself doesn't appear in ubuntu until xenial. also cryptography seems to directly build and link against libssl rather than depending on python-openssl, so seems like your patch is the only way for now unless we're dropping support for trusty | 14:26 |
| fungi | but we should remember to revisit the breakdown of system packages and python packages in that puppet module regularly | 14:27 |
| *** chandankumar is now known as chandan|sleeping | 14:28 | |
| *** e0ne has quit IRC | 14:28 | |
| *** bnemec is now known as beekneemech | 14:30 | |
| *** cody-somerville has joined #openstack-infra | 14:31 | |
| *** e0ne has joined #openstack-infra | 14:34 | |
| *** gomarivera has quit IRC | 14:35 | |
| *** stevemar__ has quit IRC | 14:36 | |
| *** gardlt has joined #openstack-infra | 14:36 | |
| *** mhickey has joined #openstack-infra | 14:37 | |
| *** stevemar__ has joined #openstack-infra | 14:37 | |
| *** qba73 has quit IRC | 14:37 | |
| *** gardlt has quit IRC | 14:38 | |
| *** gardlt has joined #openstack-infra | 14:38 | |
| *** xarses_ has joined #openstack-infra | 14:39 | |
| *** marst has joined #openstack-infra | 14:39 | |
| *** felipemonteiro__ has quit IRC | 14:41 | |
| *** matrohon has quit IRC | 14:41 | |
| *** felipemonteiro__ has joined #openstack-infra | 14:41 | |
| *** stevemar__ has quit IRC | 14:41 | |
| jeblair | fungi, pabelanger: we used to have all zuul deps from distro. i'd like to get there again eventually. i'm assuming we'll want ansible from source for a while though. | 14:42 |
| *** jamesmca_ has quit IRC | 14:42 | |
| *** jamesmca_ has joined #openstack-infra | 14:42 | |
| *** smarcet has joined #openstack-infra | 14:43 | |
| fungi | jeblair: i agree, it's just tech debt we need to carry for now and remember to revisit periodically to see what can be refactored back toward distro packages | 14:43 |
| *** xarses_ has quit IRC | 14:44 | |
| *** xarses_ has joined #openstack-infra | 14:45 | |
| *** andreww has joined #openstack-infra | 14:46 | |
| *** xarses_ has quit IRC | 14:50 | |
| *** andreww has quit IRC | 14:50 | |
| *** xarses_ has joined #openstack-infra | 14:51 | |
| *** andreww has joined #openstack-infra | 14:52 | |
| openstackgerrit | Merged openstack-infra/puppet-zuul master: Add ansible dependencies for zuulv3 https://review.openstack.org/470072 | 14:52 |
| *** Hal has quit IRC | 14:53 | |
| *** xarses_ has quit IRC | 14:55 | |
| *** xarses_ has joined #openstack-infra | 14:56 | |
| *** andreww has quit IRC | 14:56 | |
| *** andreww has joined #openstack-infra | 14:57 | |
| *** cshastri has quit IRC | 14:58 | |
| openstackgerrit | wes hayutin proposed openstack-infra/project-config master: WIP: add a multinode scenario job to check-tripleo https://review.openstack.org/470319 | 14:59 |
| *** mrmartin has joined #openstack-infra | 14:59 | |
| *** esikachev has quit IRC | 15:00 | |
| *** xarses_ has quit IRC | 15:01 | |
| *** xarses_ has joined #openstack-infra | 15:01 | |
| *** andreww has quit IRC | 15:01 | |
| openstackgerrit | Derek Higgins proposed openstack-infra/project-config master: WIP: Add job to test ironic boot from volume https://review.openstack.org/470307 | 15:03 |
| *** esikachev has joined #openstack-infra | 15:04 | |
| *** rcernin has quit IRC | 15:05 | |
| openstackgerrit | John Trowbridge proposed openstack-infra/project-config master: Add scenario coverage to quickstart gates https://review.openstack.org/470324 | 15:08 |
| openstackgerrit | Evgeny Antyshev proposed openstack-infra/puppet-pip master: Don't install pip3 if no python3 installed https://review.openstack.org/470289 | 15:08 |
| *** cshen has quit IRC | 15:09 | |
| *** xarses_ has quit IRC | 15:10 | |
| *** Hal has joined #openstack-infra | 15:11 | |
| *** dimak_ has joined #openstack-infra | 15:12 | |
| *** arahal_ has quit IRC | 15:13 | |
| *** arahal_ has joined #openstack-infra | 15:13 | |
| *** neiljerram has joined #openstack-infra | 15:13 | |
| *** gardlt has quit IRC | 15:15 | |
| *** armax has joined #openstack-infra | 15:15 | |
| *** jaypipes is now known as leakypipes | 15:16 | |
| *** jamesmca_ has quit IRC | 15:17 | |
| *** zz_dimtruck is now known as dimtruck | 15:17 | |
| pabelanger | jeblair: fungi | 15:20 |
| pabelanger | thanks, attempting to launch zuulv3.o.o again | 15:20 |
| pabelanger | sorry for multi line | 15:20 |
| *** gardlt has joined #openstack-infra | 15:22 | |
| *** bobh has joined #openstack-infra | 15:22 | |
| *** vponomaryov has quit IRC | 15:23 | |
| *** esikachev has quit IRC | 15:24 | |
| *** gardlt has quit IRC | 15:27 | |
| *** felipemonteiro__ has quit IRC | 15:27 | |
| *** gyee has joined #openstack-infra | 15:27 | |
| *** dims has quit IRC | 15:28 | |
| *** ykarel has quit IRC | 15:28 | |
| *** sambetts is now known as sambetts|afk | 15:29 | |
| *** cody-somerville has quit IRC | 15:34 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config master: Add zuulv3.o.o to cacti / zookeeper firewall https://review.openstack.org/470339 | 15:35 |
| pabelanger | jeblair: I was thinking, do we want to add status.o.o/zuulv3 for now? Rather then local UI running on zuulv3.o.o? | 15:36 |
| dhellmann | lucasagomes : yes, that patch is intended to address the failure you're seeing. Until we land and release it, we may want to just block 2.3.1: https://review.openstack.org/470340 | 15:36 |
| fungi | pabelanger: we still keep a local ui running on zuul.o.o so i don't know why that would be "instead of" | 15:36 |
| fungi | http://zuul.openstack.org/ | 15:37 |
| lucasagomes | dhellmann, fair enough! Thanks for looking into it btw | 15:37 |
| inc0 | good morning | 15:37 |
| *** hongbin has joined #openstack-infra | 15:37 | |
| inc0 | so, since voting is going well https://review.openstack.org/#/c/469265/ I might as well start doing heavy lifting and writing some code for pushing images:) | 15:38 |
| jeblair | pabelanger, fungi: yeah, let's ignore status.o.o for now | 15:38 |
| pabelanger | jeblair: fungi: okay, I'll propose patches to enable that | 15:39 |
| *** hashar is now known as hasharAway | 15:39 | |
| pabelanger | but zuulv3.o.o is online | 15:39 |
| inc0 | can you help me kick it off? what I'll be trying to do is 1. run Kolla build job http://logs.openstack.org/98/470298/1/check/gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial/50713f1/ like this | 15:39 |
| pabelanger | and installed under python3 | 15:39 |
| pabelanger | just auditing zuul.conf right now, we likely need to plumb through settings properly | 15:40 |
| inc0 | if it succeed, make a tarball and publish it somewhere temporarily | 15:40 |
| inc0 | run all of deployment jobs | 15:40 |
| inc0 | if all of them succeed, publish tarball to tarballs.o.o | 15:40 |
| inc0 | if some of them fail, delete temp tarball | 15:41 |
| pabelanger | inc0: see: http://git.openstack.org/cgit/openstack/dox/tree/tools/tarball.sh that's basically the first 1/2 for publishing we'll likely do | 15:41 |
| inc0 | and do this once a day | 15:41 |
| *** Guest75163 has quit IRC | 15:41 | |
| *** Hal has quit IRC | 15:41 | |
| inc0 | pabelanger: you don't want docker savew | 15:41 |
| inc0 | it will eat up a lot of space | 15:42 |
| pabelanger | already have 2nd 1/2 written to download onto secure node and JJB settings but waiting to see results from goverenance | 15:42 |
| fungi | the big missing piece is we don't have a tarballs.o.o publisher which deletes things right now. we expect you to just overwrite the files which are there instead | 15:42 |
| pabelanger | inc0: do you want layers? | 15:42 |
| pabelanger | if not it can be changed to docker export IIRC | 15:42 |
| inc0 | fungi: thing is, I want tarbals.o.o to have only thigns that passed deploy gates | 15:42 |
| inc0 | pabelanger: this registry hack we did is most disk space optimal thing I managed to achieve | 15:43 |
| fungi | inc0: right, we have some missing pieces for that idea | 15:43 |
| clarkb | inc0: fungi or just use a test/staging area like we do for gerrit and a published area | 15:43 |
| pabelanger | inc0: right, we don't want to be passing around the concents of a registry, we;d either want docker save / docker export | 15:43 |
| clarkb | overwrite in test/staging, copy if passes to published | 15:43 |
| inc0 | pabelanger: why? | 15:44 |
| fungi | clarkb: yep, that would likely work so long as what goes in the staging area gets overwritten with the same filename each time so we don't build up cruft in there | 15:44 |
| inc0 | that works | 15:44 |
| pabelanger | inc0: docker save / export does the same thing and does not require you to run a registry to use the images when you download them | 15:45 |
| inc0 | then we just need to ensure that there won't be race conditions between starting a deploy gate and finishing upload to staging | 15:45 |
| pabelanger | we can just docker import (Ithink) < tarball | 15:45 |
| inc0 | pabelanger: and eats up 8 times more disk space | 15:45 |
| inc0 | also takes more time | 15:45 |
| pabelanger | why would it use more disk space? | 15:45 |
| inc0 | pabelanger: not sure, but I tested that | 15:45 |
| inc0 | well | 15:46 |
| inc0 | layers are flettened | 15:46 |
| pabelanger | right | 15:46 |
| inc0 | so there is lots of not-sharing | 15:46 |
| inc0 | and we have 200+ images | 15:46 |
| pabelanger | do you want layers or no layers | 15:46 |
| fungi | the other challenge is how to engineer the job which publishes the staging artifact to the directory where we serve the known working version. it sounds like the desire is to have multiple jobs consume the staging version and only if all of them succeed make that an official copy, but right now zuulv2 doesn't have the dependency model we'd need for many to one unless they all get serialized | 15:46 |
| inc0 | so sharing is actually useful | 15:46 |
| pabelanger | docker export will flatten layers into a image blob, that is possible we can use | 15:46 |
| inc0 | docker save does flattening too, just I assume tar mechanism compresses whole lot of it | 15:47 |
| inc0 | that's why we're looking at 15 gig not 200gig | 15:47 |
| pabelanger | no, docker save keeps layers | 15:47 |
| inc0 | registry tarball is 2gig | 15:47 |
| pabelanger | https://docs.docker.com/engine/reference/commandline/save/ | 15:47 |
| inc0 | still, I tested it | 15:47 |
| pabelanger | so, you have answered my question yet. Does kolla need to keep layers or have we flatten them. Because, that will save disk space | 15:48 |
| inc0 | it might keep layers but it doesn't share them between cotnainers | 15:48 |
| inc0 | we keep layers | 15:48 |
| pabelanger | yes, it will share, you need to expore images in single command | 15:48 |
| inc0 | that will save disk space *and* a lot of uplink during pushes | 15:48 |
| inc0 | pabelanger: I did | 15:49 |
| inc0 | that's why it's *only* 15gig | 15:49 |
| *** Guest53691 is now known as rook | 15:49 | |
| inc0 | again, registry is little over 2 | 15:49 |
| inc0 | I don't really care for space if you don't care for it | 15:49 |
| inc0 | docker save is easier to manage | 15:49 |
| *** rook is now known as Guest94961 | 15:49 | |
| inc0 | I just assumed we care:) | 15:49 |
| *** matrohon has joined #openstack-infra | 15:49 | |
| *** Guest94961 is now known as rook- | 15:49 | |
| pabelanger | yes, docker save requires less infrastruture | 15:49 |
| *** rook- is now known as rook-- | 15:49 | |
| inc0 | so, lots more space that's gonna be transferred to every gate job in Kolla is ok? | 15:50 |
| pabelanger | we need to audit it I believe | 15:50 |
| inc0 | keep in mind, we're taking this tarball and use it with gates | 15:50 |
| inc0 | pabelanger: alternatively forget about registry and let us handle it | 15:50 |
| pabelanger | but compress docker save / docker registery should be the same size. It is the same data | 15:51 |
| inc0 | we can do pushing from kolla gate jobs which already exists | 15:51 |
| pabelanger | publish where? | 15:51 |
| *** Apoorva has joined #openstack-infra | 15:51 | |
| inc0 | can you add secret to gate job? | 15:51 |
| *** esikachev has joined #openstack-infra | 15:52 | |
| pabelanger | not today, but we wouldn't likely do that for zuulv3 | 15:52 |
| openstackgerrit | Doug Hellmann proposed openstack-infra/release-tools master: update semver module for python 3.5 https://review.openstack.org/470350 | 15:52 |
| pabelanger | we don't want untrusted nodes publishing artifacts directly, since anybody has root access on the node | 15:52 |
| inc0 | pabelanger: another thing | 15:52 |
| inc0 | I assume you're running docker if you're doing docker save? | 15:53 |
| inc0 | if so, running docker registry is one command | 15:53 |
| inc0 | docker run -d registry:2 | 15:53 |
| inc0 | sorry docker run -d -p5000:5000 registry:2 | 15:53 |
| inc0 | since it's very short lived registry we don't need to care any more | 15:53 |
| *** flepied has quit IRC | 15:54 | |
| *** ralonsoh has quit IRC | 15:54 | |
| pabelanger | on the trusted node, I think we'll just use docker import / docker push, since registry might now be local | 15:55 |
| pabelanger | that will likely live on control plane server and mirror to docker hub, or push directly. We need to spec that step out | 15:55 |
| openstackgerrit | Doug Hellmann proposed openstack-infra/project-config master: add python 3.5 unit test job for release-tools repository https://review.openstack.org/470352 | 15:55 |
| pabelanger | but, we shouldn't need to run a regitry between untrusted / trusted nodes | 15:55 |
| inc0 | I don't want that either | 15:56 |
| pabelanger | we can use docker import / export commands to transfer images, and compress | 15:56 |
| inc0 | so here's my plan | 15:56 |
| *** eantyshev has joined #openstack-infra | 15:56 | |
| inc0 | we handle tarballs.o.o like I described above | 15:56 |
| inc0 | so in tarballs.o.o/kolla/images we have fresh and tested registry tarball | 15:56 |
| *** flepied has joined #openstack-infra | 15:56 | |
| inc0 | on secure node once a day we run, in cron, this | 15:56 |
| *** salv-orl_ has quit IRC | 15:57 | |
| inc0 | https://github.com/inc0/kolla_dockerhub_pusher | 15:57 |
| *** eantyshev has left #openstack-infra | 15:57 | |
| fungi | it would be good to converge on a similar model to what mnaser is trying for in his dockerhub publisher spec https://review.openstack.org/447524 | 15:57 |
| inc0 | yes, I've already written tool that will do downloading tarball, running local registry, changing tags and pushing to hub | 15:57 |
| pabelanger | inc0: we don't build on trusted nodes, that happend on untrusted. We only publish from trusted today | 15:57 |
| fungi | at least to be able to reuse some of the same automation and maybe servers | 15:57 |
| *** rkukura has quit IRC | 15:57 | |
| pabelanger | inc0: but yes, it can be periodic | 15:58 |
| inc0 | pabelanger: no building on trusted nodes | 15:58 |
| *** dhinesh has joined #openstack-infra | 15:58 | |
| *** stevemar__ has joined #openstack-infra | 15:59 | |
| inc0 | fungi: what mnaser wrote is exactly what I described | 15:59 |
| *** jascott1 has joined #openstack-infra | 15:59 | |
| pabelanger | inc0: once we get the JJB configuration up for review, I will CC you. I am pretty sure it will work the way you are wanting | 15:59 |
| fungi | yep, saying let's get that spec polished up and get some consensus on it | 15:59 |
| fungi | otherwise we risk implementing something which is at odds with the spec we agree on | 16:00 |
| pabelanger | fungi: yes, that is what I am holding off on. Until goverance approves their spec, we are blocked | 16:00 |
| *** tesseract has quit IRC | 16:00 | |
| pabelanger | but I see this working the same way we do python tarballs today, just with a few docker commands instead | 16:00 |
| *** dimak_ has quit IRC | 16:01 | |
| fungi | pabelanger: the tc resolution does need to come before the dockerhub publisher spec, but it seems like we can at least address some of the concerns called out in our publisher spec in parallel | 16:01 |
| inc0 | also pipeline to saving to tarballs isn't blocked by TC spec and tha'ts gonna be a lot of work | 16:01 |
| *** arahal__ has joined #openstack-infra | 16:02 | |
| inc0 | fungi: anyway, I'm ok with this spec just to be clear | 16:02 |
| pabelanger | yes, I have been doing POC with dox, be it has some images today and already stuff on docker hub | 16:02 |
| *** cody-somerville has joined #openstack-infra | 16:02 | |
| *** cody-somerville has joined #openstack-infra | 16:02 | |
| pabelanger | the big question is if we publish directly to docker hub | 16:02 |
| inc0 | how else? | 16:03 |
| pabelanger | or to a local registry in control plan and have it mirror | 16:03 |
| inc0 | as in registry.openstack.org? | 16:03 |
| pabelanger | something along those lines | 16:03 |
| inc0 | it's cool, but I don't think there is a lot of value in it | 16:03 |
| inc0 | just more work for us | 16:03 |
| fungi | exposed as http://mirror.iad.rax.openstack.org/docker-registry/ et al | 16:04 |
| pabelanger | right | 16:04 |
| *** esikachev has quit IRC | 16:04 | |
| inc0 | I wanted that before we came up with dockerhub+cache idea:) | 16:04 |
| pabelanger | the value is, trusted node is local hope to registry, so uploads should be fast, then mirror can happen over lower speed links | 16:04 |
| inc0 | now I like dockerhub+cache more | 16:04 |
| fungi | i have no preference between distributed registry mirror and cache of centralized mirror | 16:05 |
| *** aarefiev_afk has quit IRC | 16:05 | |
| pabelanger | I suspect we'd try to publish directly first, we if get network timeouts then likely do the local registry mirror approach | 16:05 |
| *** arahal_ has quit IRC | 16:05 | |
| inc0 | that too, we don't need to distribute local registry across nodepools | 16:05 |
| fungi | the latter may also perform better... dragging very large files through afs is not great from a performance standpoint | 16:05 |
| pabelanger | inc0: we'll we likely want to setup a reverse proxy cache, if you are saying your images are large | 16:06 |
| inc0 | fungi: if we mirror registry that's not little files | 16:06 |
| inc0 | that's a lot of them | 16:06 |
| clarkb | pabelanger: we already have that though right? | 16:06 |
| *** matrohon has quit IRC | 16:06 | |
| pabelanger | clarkb: yup, jobs just need to opt into them | 16:06 |
| *** gardlt has joined #openstack-infra | 16:07 | |
| pabelanger | clarkb: but sounds like we might need to bump the cache size | 16:07 |
| inc0 | our images aren't that large, as I said, tarball of registry, which we can always use, is less than 3 gig | 16:07 |
| fungi | inc0: right, that's what i'm saying. putting docker images in afs is likely to perform poorly due to file size and churn frequency | 16:07 |
| inc0 | cache was about images we don't build, like ubuntu:16.04 | 16:07 |
| fungi | 3gb is a pretty large file for a network filesystem written in the 80s and implemented over udp | 16:07 |
| inc0 | we already upload tarballs and use it | 16:07 |
| pabelanger | inc0: are you using NODEPOOL_DOCKER_REGISTRY_PROXY today for jobs? | 16:07 |
| pabelanger | eg: http://mirror.regionone.infracloud-vanilla.openstack.org:8081/registry-1.docker | 16:08 |
| *** gaurangt_ has quit IRC | 16:08 | |
| inc0 | I think so, let me confirm that for you | 16:08 |
| *** smarcet has quit IRC | 16:08 | |
| *** dhinesh has quit IRC | 16:09 | |
| pabelanger | fungi: clarkb: jeblair: do you mind a review of https://review.openstack.org/#/c/470339/ needed for zuulv3.o.o to access zookeeper | 16:10 |
| *** gaurangt has joined #openstack-infra | 16:11 | |
| *** mriedem is now known as mriedem_burgerti | 16:11 | |
| *** mriedem_burgerti is now known as mriedem | 16:11 | |
| *** felipemonteiro has joined #openstack-infra | 16:12 | |
| *** amoralej is now known as amoralej|off | 16:12 | |
| inc0 | kolla doesn't - but kolla only downloads base image so that's not that horrible | 16:12 |
| inc0 | will fix tho | 16:12 |
| inc0 | kolla-k8s which was reason we care, does | 16:13 |
| *** felipemonteiro__ has joined #openstack-infra | 16:13 | |
| *** e0ne_ has joined #openstack-infra | 16:13 | |
| *** yamamoto has quit IRC | 16:14 | |
| *** e0ne has quit IRC | 16:15 | |
| *** yamamoto has joined #openstack-infra | 16:15 | |
| *** pcaruana has quit IRC | 16:16 | |
| *** felipemonteiro has quit IRC | 16:16 | |
| *** dimak_ has joined #openstack-infra | 16:17 | |
| openstackgerrit | Tom Barron proposed openstack-infra/project-config master: remove manila mitaka lvm job https://review.openstack.org/470358 | 16:17 |
| *** yamamoto has quit IRC | 16:19 | |
| *** e0ne_ has quit IRC | 16:21 | |
| *** dimak_ has quit IRC | 16:21 | |
| openstackgerrit | Merged openstack-infra/system-config master: Add zuulv3.o.o to cacti / zookeeper firewall https://review.openstack.org/470339 | 16:23 |
| *** ericyoung has quit IRC | 16:29 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-zuul master: Add zuulv3 configuration file https://review.openstack.org/470362 | 16:29 |
| *** hongbin has quit IRC | 16:29 | |
| *** mdrabe has quit IRC | 16:30 | |
| *** chandan|sleeping is now known as chandankumar | 16:30 | |
| *** jaosorior has quit IRC | 16:30 | |
| inc0 | ok, so, back to inital question, how can I make job that will run daily? | 16:31 |
| inc0 | + have some staging/temporary publisher space | 16:31 |
| *** gomarivera has joined #openstack-infra | 16:32 | |
| fungi | define a periodic job to do the build, and give it a tarballs.o.o publisher which writes the staging version of the file to a predictable name and/or path that indicates it's the staging version and not a known working one | 16:33 |
| fungi | then you can have a (series of?) jobs depending on the success of that job to test the staging copy and, if they also succeed, publish it to another path on tarballs.o.o for a known-good artifact | 16:34 |
| *** knangia_ has joined #openstack-infra | 16:35 | |
| *** brd_ has quit IRC | 16:35 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config master: Use zuulv3 syntax for configuration https://review.openstack.org/470363 | 16:35 |
| inc0 | also we'll need to figure out how to include ids of jobs that images passed, but that's for later | 16:35 |
| fungi | this results in two artifacts on tarballs.o.o... a staging artifact which may or may not be usable, and a known-working artifact which has been confirmed to pass in your jobs | 16:36 |
| pabelanger | jeblair: do you mind ensure I setup the proper settings for zuul.conf ^. There is a depends-on which creates a new zuulv3.conf.erb file. | 16:36 |
| *** xarses_ has joined #openstack-infra | 16:37 | |
| *** smarcet has joined #openstack-infra | 16:37 | |
| fungi | inc0: i suppose your validation jobs could also replace another file alongside the known working one which contains a list of the jobs that ran against it | 16:37 |
| fungi | just a small text file maybe | 16:37 |
| inc0 | yeah we'll figure it out | 16:37 |
| *** Swami has joined #openstack-infra | 16:37 | |
| inc0 | do you have expample of periodic job anywahere? | 16:38 |
| fungi | maybe append to a similar file alongside the staging copy, and then the final blessing job republishes the list of jobs which ran alongside the known-working artifact | 16:38 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Consider hostname in determining local builds https://review.openstack.org/470364 | 16:39 |
| fungi | inc0: periodic jobs are basically defined like any other ci job, they just need to be aware that they won't receive envvars from zuul identifying a change id or branch name since those aren't contextual to a periodic job and therefore need to be set within the job if it will need them | 16:39 |
| fungi | inc0: here are some, for example http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/python-bitrot-jobs.yaml | 16:40 |
| inc0 | what about getting to logs and such? | 16:40 |
| *** jpich has quit IRC | 16:42 | |
| fungi | inc0: the logs get published to http://logs.openstack.org/periodic/<jobname>/ but if you write them such that they emit subunit data then they can also be tracked more conveniently at http://status.openstack.org/openstack-health/ or by querying the http://health.openstack.org/ api server | 16:42 |
| inc0 | yeah I'll do both, cool | 16:43 |
| inc0 | last question, can I manually trigger a job? | 16:43 |
| inc0 | well probably not last, but last one I have now;) | 16:43 |
| *** mdrabe has joined #openstack-infra | 16:44 | |
| fungi | not a periodic job, no... though you could write a basically equivalent job for the experimental pipeline and run it against a dummy change by leaving a "check experimental" comment | 16:44 |
| inc0 | hmm...right, I'm thinking how to fix it if something goes wrong | 16:45 |
| inc0 | since it's gonna be probably one of more complex periodic job pipeline out there | 16:45 |
| fungi | you can request that a zuul admin retrigger a job for you | 16:45 |
| fungi | i'm not sure we have the plumbing we need to do that on periodic pipelines though | 16:46 |
| *** rwsu has joined #openstack-infra | 16:46 | |
| inc0 | well probably something we'll need to work on then, but later I guess | 16:46 |
| fungi | but the experimental pipeline job would at least give you a means to iterate more quickly on it at first until you get it to the point where it's doing what you expect | 16:47 |
| fungi | and then probably you'd remove the experimental pipeline job from that point on so that random people can't force an update | 16:47 |
| inc0 | yeah, that's what I'll do right away, we can start thinking of periodic manual trigger in the meantime | 16:47 |
| inc0 | well...on the flip side, there is no risk of having too much updates on this besides ddos;) | 16:48 |
| *** eharney has quit IRC | 16:48 | |
| *** florianf has quit IRC | 16:48 | |
| openstackgerrit | Slawek Kaplonski proposed openstack-infra/shade master: Remove designate client from shade's dependencies https://review.openstack.org/470366 | 16:49 |
| *** trown is now known as trown|lunch | 16:51 | |
| *** derekh has quit IRC | 16:51 | |
| *** claudiub has quit IRC | 16:52 | |
| fungi | inc0: the biggest risk i can see is races since we can't guarantee multiple experimental jobs won't run in parallel so you _could_ end up blessing a different artifact than you tested (as opposed to a series of periodics job where we know they only run once in a given timeframe) | 16:54 |
| inc0 | good point | 16:54 |
| inc0 | anyway, off to zuul conf I go to create initial build jobs | 16:55 |
| fungi | have fun storming the castle | 16:55 |
| *** Guest75163 has joined #openstack-infra | 16:57 | |
| *** dtantsur is now known as dtantsur|afk | 16:58 | |
| jeblair | pabelanger: that looks fine. theoretically, the same template should work for both, but it may be easier to have two for now. | 16:58 |
| jeblair | pabelanger: regarding 470363, you put the connections object in hiera, but i don't see it added to public hiera | 17:00 |
| jeblair | pabelanger: were you planning on adding the whole thing to private hiera? | 17:00 |
| *** huanxie has quit IRC | 17:03 | |
| *** huanxie has joined #openstack-infra | 17:03 | |
| *** rook-- is now known as rook | 17:03 | |
| *** eharney has joined #openstack-infra | 17:03 | |
| *** spzala has quit IRC | 17:04 | |
| pabelanger | jeblair: Ya, that was based on how we did sites / nodes today. But I could move it directly into puppet, for each connection. I was thinking yaml format might be easier to maintain then in puppet | 17:05 |
| pabelanger | I guess public hiera should be okay? | 17:05 |
| jeblair | pabelanger: there isn't anything secret about it, except for github api key | 17:05 |
| *** cody-somerville has quit IRC | 17:06 | |
| pabelanger | k, let me update the patch | 17:06 |
| jeblair | i don't suppose we can mix the two... | 17:06 |
| pabelanger | I am not sure to be honest | 17:06 |
| *** cody-somerville has joined #openstack-infra | 17:06 | |
| jeblair | if we can't, then i reluctantly think just a puppet data structure is the best option | 17:06 |
| jeblair | that way everything can be public, and we can get the github api key from hiera | 17:07 |
| jeblair | (private hiera) | 17:07 |
| pabelanger | k, let me work on that | 17:07 |
| jeblair | if there's a way to have a public hiera value reference a private hiera value, that would be ideal. | 17:07 |
| pabelanger | EmilienM: ^ do you happen to know if heira works like that? | 17:08 |
| *** dhajare-brb has quit IRC | 17:09 | |
| jeblair | https://docs.puppet.com/hiera/3.3/variables.html#the-hiera-lookup-function | 17:10 |
| *** dims has joined #openstack-infra | 17:10 | |
| jeblair | maybe like that ^? | 17:10 |
| pabelanger | we can try | 17:10 |
| pabelanger | do we have a github API key atm? | 17:11 |
| *** brd has joined #openstack-infra | 17:11 | |
| clarkb | we have the account that closes pull requests but I think that may be shared with the project setup account. Might be ebtter to make a new smaller scoped token for that | 17:11 |
| jeblair | i think mordred started setting up a github 'app' | 17:12 |
| mordred | yes. I have set up a github "app" | 17:13 |
| jeblair | i'm not sure i know exactly what we need to put in the config file yet though | 17:13 |
| mordred | I will work next on putting info from it into passwords and hiera | 17:13 |
| jeblair | i don't think we need to set up the github connection right now; mostly just want to prepare for it. | 17:13 |
| *** Guest75163 has quit IRC | 17:13 | |
| mordred | ++ | 17:13 |
| *** Guest75163 has joined #openstack-infra | 17:14 | |
| jeblair | (like, i expect us to do so on the order of about a week, so we should design for it now) | 17:14 |
| pabelanger | pass: "%{hiera('cloudapidocs_ftp_password')}" | 17:14 |
| pabelanger | jeblair: what we do today | 17:14 |
| pabelanger | so it must work | 17:14 |
| jeblair | pabelanger: oh hey cool | 17:14 |
| jeblair | pabelanger: so let's do that. that'll be perfect. | 17:14 |
| jeblair | heh, and that's from zuul :) | 17:15 |
| jeblair | pabelanger: so i think we can stick the whole connection structure in public hiera, and we'll know that later when we add a github connection to it, we'll be able to do that. | 17:15 |
| *** yamamoto has joined #openstack-infra | 17:16 | |
| jeblair | pabelanger: i think we should add the public hiera entry in change 470363 | 17:16 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config master: Use zuulv3 syntax for configuration https://review.openstack.org/470363 | 17:17 |
| pabelanger | jeblair: agree | 17:17 |
| pabelanger | that's what we have on zuulv3-dev.o.o today | 17:17 |
| *** jamesmca_ has joined #openstack-infra | 17:17 | |
| *** xyang1 has joined #openstack-infra | 17:17 | |
| *** yamamoto has quit IRC | 17:21 | |
| openstackgerrit | Colleen Murphy proposed openstack-infra/project-config master: Test keystone with newly split tempest plugin https://review.openstack.org/470373 | 17:21 |
| *** jamesmca_ has quit IRC | 17:21 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid-resources master: added new public endpoint https://review.openstack.org/470374 | 17:22 |
| *** mhickey has quit IRC | 17:22 | |
| *** esberglu has quit IRC | 17:24 | |
| openstackgerrit | Merged openstack-infra/openstackid-resources master: added new public endpoint https://review.openstack.org/470374 | 17:24 |
| *** dims has quit IRC | 17:24 | |
| *** esberglu has joined #openstack-infra | 17:24 | |
| *** dims has joined #openstack-infra | 17:26 | |
| *** harlowja has quit IRC | 17:28 | |
| *** Guest75163 has quit IRC | 17:28 | |
| *** esberglu has quit IRC | 17:29 | |
| *** funzo has quit IRC | 17:29 | |
| *** florianf has joined #openstack-infra | 17:29 | |
| *** funzo has joined #openstack-infra | 17:29 | |
| *** funzo has quit IRC | 17:29 | |
| *** lucasagomes is now known as lucas-afk | 17:30 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid-resources master: hotfix test https://review.openstack.org/470375 | 17:32 |
| pabelanger | clarkb: fungi: do you mind a +3 on 470363 and depends-on | 17:32 |
| openstackgerrit | Merged openstack-infra/openstackid-resources master: hotfix test https://review.openstack.org/470375 | 17:33 |
| pabelanger | fungi: thank you | 17:36 |
| fungi | pabelanger: seems fine to me but i had a question (inline) before approving | 17:37 |
| *** esberglu has joined #openstack-infra | 17:37 | |
| pabelanger | fungi: ah, yes. that can be removed | 17:38 |
| pabelanger | let me fix quickly | 17:38 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config master: Use zuulv3 syntax for configuration https://review.openstack.org/470363 | 17:39 |
| openstackgerrit | Merged openstack-infra/puppet-zuul master: Add zuulv3 configuration file https://review.openstack.org/470362 | 17:40 |
| *** gomarivera has quit IRC | 17:41 | |
| openstackgerrit | Colleen Murphy proposed openstack-infra/project-config master: Test keystone with newly split tempest plugin https://review.openstack.org/470373 | 17:41 |
| *** dhinesh has joined #openstack-infra | 17:45 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_post_checkout https://review.openstack.org/470381 | 17:47 |
| *** rbrndt has quit IRC | 17:48 | |
| *** jkilpatr has quit IRC | 17:48 | |
| *** lin_yang has joined #openstack-infra | 17:50 | |
| *** cody-somerville has quit IRC | 17:51 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/devstack-gate master: DNM reduce service set for swift func testing https://review.openstack.org/470383 | 17:51 |
| *** neiljerram has quit IRC | 17:52 | |
| pabelanger | fungi: Oh, heh, Our testing failes for the change now | 17:52 |
| pabelanger | I guess that is why we defaulted to [] | 17:52 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config master: Use zuulv3 syntax for configuration https://review.openstack.org/470363 | 17:53 |
| fungi | pabelanger: it shouldn't unless there's a problem with the tests | 17:53 |
| pabelanger | fungi: yes, our tests just make heira data strings | 17:53 |
| fungi | pabelanger: oh, i see. we need some means to specify a different data type in that case | 17:54 |
| fungi | that is indeed a challenge in a language that doesn't separate type declaration | 17:54 |
| pabelanger | ya, this is how we get zuul_site to work today, just default to [] | 17:55 |
| EmilienM | pabelanger: it's evening for me I'll look later this weekend. On my phone now sorry... | 17:56 |
| pabelanger | EmilienM: np, we figured it out. Thanks for replying | 17:57 |
| *** e0ne has joined #openstack-infra | 17:57 | |
| *** csomerville has joined #openstack-infra | 17:59 | |
| *** florianf has quit IRC | 18:00 | |
| *** Apoorva_ has joined #openstack-infra | 18:00 | |
| *** spzala has joined #openstack-infra | 18:02 | |
| *** Apoorva has quit IRC | 18:03 | |
| *** jkilpatr has joined #openstack-infra | 18:03 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_post_and_master_checkout https://review.openstack.org/470385 | 18:05 |
| *** jpena is now known as jpena|off | 18:05 | |
| *** trown|lunch is now known as trown | 18:06 | |
| *** lbragstad has quit IRC | 18:08 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/system-config master: Updated OpenStackId dev instance assets URL https://review.openstack.org/470386 | 18:08 |
| *** yamahata has joined #openstack-infra | 18:08 | |
| *** cody-somerville has joined #openstack-infra | 18:11 | |
| fungi | clarkb: ianychoi: ^ is related to what we were discussing yesterday | 18:14 |
| *** AJaeger has quit IRC | 18:14 | |
| *** csomerville has quit IRC | 18:14 | |
| clarkb | fungi: cool, but thee is still the basic auth problem? or was I msireading that that was an issue? | 18:15 |
| *** dims has quit IRC | 18:16 | |
| openstackgerrit | Waldemar Znoinski proposed openstack-infra/glean master: Revise systemd determination to verify systemctl presence https://review.openstack.org/279818 | 18:16 |
| fungi | clarkb: smarcet either already confirmed or was going to adjust it so that the password reset and profile urls weren't behind basic auth any longer | 18:16 |
| smarcet | fungi: clarkb: i will adjust that, i mean, still working on that | 18:17 |
| clarkb | cool | 18:17 |
| *** cody-somerville has quit IRC | 18:17 | |
| *** dims has joined #openstack-infra | 18:17 | |
| *** yamamoto has joined #openstack-infra | 18:17 | |
| fungi | if teh password reset link on openstackid-dev.o.o isn't able to be followed for a bit after that merges it's not really any worse than when it was going to the wrong place | 18:17 |
| openstackgerrit | Waldemar Znoinski proposed openstack-infra/glean master: Revise systemd determination to verify systemctl presence https://review.openstack.org/279818 | 18:18 |
| *** cody-somerville has joined #openstack-infra | 18:18 | |
| *** cody-somerville has joined #openstack-infra | 18:18 | |
| *** csomerville has joined #openstack-infra | 18:20 | |
| smarcet | fungi: yes the link will be available, i will remove basic auth | 18:20 |
| smarcet | now working on properly handling those emails | 18:20 |
| weshay | pabelanger, sorry I always ping you for infra reviews.. but a guys gotta make a livin some how :) https://review.openstack.org/#/c/470324/ | 18:21 |
| *** yamamoto has quit IRC | 18:23 | |
| *** cody-somerville has quit IRC | 18:23 | |
| *** rbrndt has joined #openstack-infra | 18:26 | |
| *** funzo has joined #openstack-infra | 18:30 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_git_cache_dir https://review.openstack.org/470389 | 18:32 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_upgrade https://review.openstack.org/470390 | 18:32 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_periodic_update https://review.openstack.org/470391 | 18:32 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_post_checkout https://review.openstack.org/470392 | 18:32 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Rename TestCloner TestExecutorRepos https://review.openstack.org/470393 | 18:32 |
| *** mugsie has quit IRC | 18:32 | |
| *** cody-somerville has joined #openstack-infra | 18:33 | |
| *** wolverineav has quit IRC | 18:34 | |
| *** wolverineav has joined #openstack-infra | 18:34 | |
| *** funzo has quit IRC | 18:34 | |
| *** mugsie has joined #openstack-infra | 18:35 | |
| *** mugsie has joined #openstack-infra | 18:35 | |
| *** csomerville has quit IRC | 18:35 | |
| *** gomarivera has joined #openstack-infra | 18:36 | |
| *** spzala has quit IRC | 18:37 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_post_checkout https://review.openstack.org/470392 | 18:37 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Rename TestCloner TestExecutorRepos https://review.openstack.org/470393 | 18:37 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_post_checkout https://review.openstack.org/470381 | 18:37 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_post_and_master_checkout https://review.openstack.org/470385 | 18:37 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_git_cache_dir https://review.openstack.org/470389 | 18:37 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_upgrade https://review.openstack.org/470390 | 18:38 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_periodic_update https://review.openstack.org/470391 | 18:38 |
| *** e0ne has quit IRC | 18:38 | |
| *** lbragstad has joined #openstack-infra | 18:38 | |
| *** e0ne has joined #openstack-infra | 18:38 | |
| *** jamesmca_ has joined #openstack-infra | 18:39 | |
| *** lbragstad has quit IRC | 18:39 | |
| *** asselin__ has quit IRC | 18:39 | |
| *** wolverineav has quit IRC | 18:39 | |
| *** dizquierdo has quit IRC | 18:40 | |
| *** lbragstad has joined #openstack-infra | 18:40 | |
| openstackgerrit | Merged openstack-infra/system-config master: Use zuulv3 syntax for configuration https://review.openstack.org/470363 | 18:41 |
| *** dhajare-brb has joined #openstack-infra | 18:41 | |
| openstackgerrit | Merged openstack-infra/system-config master: Updated OpenStackId dev instance assets URL https://review.openstack.org/470386 | 18:41 |
| *** lbragstad has quit IRC | 18:42 | |
| *** electrofelix has quit IRC | 18:42 | |
| *** jkilpatr has quit IRC | 18:42 | |
| *** jamesmca_ has quit IRC | 18:43 | |
| *** lbragstad has joined #openstack-infra | 18:44 | |
| *** e0ne has quit IRC | 18:45 | |
| *** lbragstad_ has joined #openstack-infra | 18:46 | |
| *** lbragstad has quit IRC | 18:46 | |
| *** lbragstad_ has quit IRC | 18:47 | |
| *** lbragstad has joined #openstack-infra | 18:48 | |
| *** lbragstad has quit IRC | 18:49 | |
| *** lbragstad__ has joined #openstack-infra | 18:49 | |
| *** lbragstad__ has quit IRC | 18:49 | |
| *** lbragstad__ has joined #openstack-infra | 18:49 | |
| smarcet | ianychoi: clarkb: fungi: patch on https://devbranch.openstack.org/ already landed, now u will be able to reset your password and re issue the verification email | 18:49 |
| pabelanger | jeblair: any objection if I stop zuulv3-dev for a few minutes. would like to see if zuulv3.o.o can establish a connection with gerrit | 18:50 |
| *** lbragstad__ has quit IRC | 18:51 | |
| *** lbragstad__ has joined #openstack-infra | 18:51 | |
| jeblair | pabelanger: why do you have to stop dev for that? | 18:53 |
| *** MasterOfBugs has joined #openstack-infra | 18:53 | |
| openstackgerrit | Felipe Monteiro proposed openstack-infra/project-config master: Use keystone uuid tokens to optimize test run time https://review.openstack.org/470399 | 18:54 |
| *** lbragstad__ has quit IRC | 18:54 | |
| jeblair | pabelanger: note that if the bwrap changes land, we won't be able to restart zuulv3-dev | 18:54 |
| *** lbragstad__ has joined #openstack-infra | 18:54 | |
| jeblair | looks like they hit a test race and haven't landed yet, but i recheck | 18:54 |
| pabelanger | jeblair: I was just concerned using the same credentials over both servers. | 18:55 |
| pabelanger | but, if you are okay with that, I can try to start zuul-scheduler on zuulv3.o.o | 18:55 |
| *** jkilpatr has joined #openstack-infra | 18:55 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-zuul master: Initial commit of zuul-executor https://review.openstack.org/469994 | 18:56 |
| pabelanger | jeblair: https://review.openstack.org/#/c/469995 should get us a zuul-executor next | 18:57 |
| pabelanger | actually, need to make sure it is using python3 | 18:57 |
| jeblair | pabelanger: yeah, should be fine to share creds | 18:57 |
| * jeblair lunches | 18:57 | |
| pabelanger | okay, I'll start it up in a moment here | 18:57 |
| *** lbragstad__ has quit IRC | 18:59 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config master: Create ze01.openstack.org https://review.openstack.org/469995 | 18:59 |
| *** gomarivera has quit IRC | 19:00 | |
| *** gtmanfred has quit IRC | 19:02 | |
| *** gtmanfred has joined #openstack-infra | 19:02 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/devstack-gate master: DNM reduce service set for swift func testing https://review.openstack.org/470383 | 19:03 |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid master: Added Favorite icon to main template https://review.openstack.org/470405 | 19:06 |
| *** eharney has quit IRC | 19:07 | |
| *** spzala has joined #openstack-infra | 19:13 | |
| *** lbragstad__ has joined #openstack-infra | 19:13 | |
| *** yland has joined #openstack-infra | 19:13 | |
| *** lbragstad__ is now known as lbragstad | 19:13 | |
| *** yland is now known as Guest54185 | 19:14 | |
| *** gyee has quit IRC | 19:14 | |
| *** jamesmca_ has joined #openstack-infra | 19:14 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/puppet-openstackid master: Updated session cookie params https://review.openstack.org/470407 | 19:16 |
| *** gyee has joined #openstack-infra | 19:17 | |
| *** spzala has quit IRC | 19:18 | |
| smarcet | fungi: could we test this https://review.openstack.org/#/c/470407/1 only on devinstance ? and stop production one ? thx ! | 19:18 |
| *** jamesmca_ has quit IRC | 19:19 | |
| openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci master: TEST: Deploy 2 ceph nodes in updates job https://review.openstack.org/470409 | 19:19 |
| *** yamamoto has joined #openstack-infra | 19:19 | |
| *** eharney has joined #openstack-infra | 19:20 | |
| fungi | smarcet: sure, i've put openstackid.org (production) in our emergency disable list temporarily. once i confirm that puppet really has stopped updating it we can go ahead there | 19:21 |
| smarcet | fungi: thx :) | 19:21 |
| openstackgerrit | Dirk Mueller proposed openstack-infra/devstack-gate master: openSUSE enablement for devstack-gate https://review.openstack.org/468685 | 19:22 |
| *** e0ne has joined #openstack-infra | 19:22 | |
| fungi | i'm being extra careful given we've had some instances in the past where we failed to actually disable puppet updates for various reasons, so any more i give it some time and make sure one pass never touches the server before i'm confident it worked | 19:23 |
| pabelanger | :( I have to rebuild zuulv3.o.o. It looks like I mistakening launched a trusty image | 19:23 |
| pabelanger | I left --image of my launch-node.py command | 19:23 |
| fungi | that'll do it | 19:23 |
| fungi | we might be approaching the point where it's time to update the default image name for that script | 19:24 |
| *** yamamoto has quit IRC | 19:24 | |
| pabelanger | agree | 19:24 |
| clarkb | dirk: looks like tempest on suse (with swift even) is raelly close now? | 19:25 |
| *** Guest54185 has quit IRC | 19:26 | |
| dirk | clarkb: it passes | 19:26 |
| dirk | clarkb: I'm just waiting for the update to be released, currently posting a experimental run using the test update | 19:26 |
| dirk | clarkb: main issue is lack of reviews :( | 19:26 |
| openstackgerrit | Merged openstack-infra/openstackid master: Added Favorite icon to main template https://review.openstack.org/470405 | 19:27 |
| clarkb | dirk: the update will include liberasurecode-devel? | 19:27 |
| clarkb | dirk: ya it hurts that sdague is on vacation as I think he is the primary reviewer of much of the devstack side | 19:27 |
| dirk | clarkb: yep see https://review.openstack.org/#/c/468676/7/tools/install_prereqs.sh | 19:27 |
| dirk | clarkb: on devstack-gate? | 19:28 |
| dirk | clarkb: e.g. the top 10 reviews of this query: https://review.openstack.org/#/q/project:openstack-infra/devstack-gate+status:open could all be reviewed and merged relatively quickly (maybe except mine :) ) | 19:28 |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-zuul master: Clean up zuul-scheduler init script https://review.openstack.org/470414 | 19:29 |
| clarkb | dirk: a good chunk of those changes depend on deciding if we want to export journal logs in native or export format | 19:30 |
| *** funzo has joined #openstack-infra | 19:31 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Abort jobs if JobDir disk usage exceeds an optional maximum https://review.openstack.org/463022 | 19:31 |
| clarkb | dirk: I personally think we should just use export format then my change merges, your change is simpler, and the change to install systemd-journal-remote earlier goes away | 19:31 |
| clarkb | dirk: and that in particular I'd like sdague input on | 19:31 |
| dirk | clarkb: true,b ut it doesn't necessarily block on that.. there are fixes that make it work (by ignoring the whole thing for suse) for now | 19:32 |
| dirk | clarkb: the addition of systemd-journal-remote is btw work in progress | 19:32 |
| dirk | clarkb: it might actually happen soon | 19:32 |
| dirk | so then it would be available for suse as well | 19:32 |
| dirk | (where soon is the matter of weeks, not days) | 19:32 |
| clarkb | thats good to know, but I also think its just a better idea to use export format since you get a single file that you download instead of multiple and avoid package installs entirely | 19:33 |
| openstackgerrit | K Jonathan Harker proposed openstack-infra/zuul feature/zuulv3: Handle job or build is None in formatUrlPattern https://review.openstack.org/470415 | 19:34 |
| *** e0ne has quit IRC | 19:34 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add SSH Agent Primitives and usage https://review.openstack.org/468173 | 19:34 |
| *** dimtruck is now known as zz_dimtruck | 19:34 | |
| dirk | clarkb: I agree.. | 19:34 |
| dirk | clarkb: nevertheless this blocks me from being able to run experimental checks for devstack reviews against suse as the devstack-gate change isn't merged.. | 19:35 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add support for bwrap https://review.openstack.org/453851 | 19:35 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 19:35 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 19:35 |
| dirk | clarkb: and it should be easy to rebase the "delete all that jremote stuff" patch on top of that patchset | 19:36 |
| clarkb | dirk: ya I have +2'd it | 19:36 |
| *** funzo has quit IRC | 19:37 | |
| dirk | clarkb: thanks. would you mind to review https://review.openstack.org/#/c/465909/ as well? it's a dependency of that patch right now | 19:37 |
| clarkb | yes I haev -1'd it | 19:38 |
| openstackgerrit | sebastian marcet proposed openstack-infra/puppet-openstackid master: Updated session cookie params https://review.openstack.org/470407 | 19:38 |
| clarkb | bceause again rather than repeatledly fixing a broken system I'd rather just remove it | 19:38 |
| clarkb | there shouldn't be a dependency with the suse enablement change because you are just using true in suse for now | 19:39 |
| *** spzala has joined #openstack-infra | 19:42 | |
| dirk | clarkb: gerrit said it would conflict.. :-) | 19:43 |
| *** hasharAway is now known as hashar | 19:43 | |
| dirk | ok, I'll rebase again without it | 19:43 |
| *** hongbin has joined #openstack-infra | 19:43 | |
| openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck master: Add query for neutron dhcpv6 bug 1655567 https://review.openstack.org/470416 | 19:43 |
| openstack | bug 1655567 in neutron "DHCPv6 failures with "address already allocated in subnet" error" [High,In progress] https://launchpad.net/bugs/1655567 - Assigned to Miguel Lavalle (minsel) | 19:43 |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid master: Updated login form https://review.openstack.org/470417 | 19:44 |
| dirk | clarkb: could you add a Closes-Bug: ... then to your change as well? | 19:44 |
| openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci master: Allow 1 ceph osd for updates job https://review.openstack.org/470418 | 19:46 |
| clarkb | I can, I didn't realize there was a bug until raeding that change | 19:46 |
| clarkb | (I had pushed in order to make life easier on suse) | 19:46 |
| openstackgerrit | Dirk Mueller proposed openstack-infra/devstack-gate master: openSUSE enablement for devstack-gate https://review.openstack.org/468685 | 19:48 |
| openstackgerrit | Matthew Thode proposed openstack-infra/project-config master: Create the requirements stable maintence team https://review.openstack.org/470419 | 19:49 |
| openstackgerrit | Clark Boylan proposed openstack-infra/devstack-gate master: Store journal in export format not native https://review.openstack.org/468705 | 19:49 |
| clarkb | dirk: ^ commit message updated | 19:49 |
| clarkb | fwiw breaking apt as part of your job run seems like the root bug (we should handle it graceufll though...) | 19:49 |
| pabelanger | okay, zuul-scheduler started | 19:51 |
| pabelanger | but waiting for gearman connection | 19:51 |
| openstackgerrit | Dirk Mueller proposed openstack-infra/devstack-gate master: openSUSE enablement for devstack-gate https://review.openstack.org/468685 | 19:51 |
| pabelanger | guess I should get ze01.o.o going now | 19:51 |
| openstackgerrit | Merged openstack-infra/openstackid master: Updated login form https://review.openstack.org/470417 | 19:52 |
| dirk | clarkb: not always easy I guess (e.g. when running out of disk space ) | 19:53 |
| pabelanger | Oh, maybe something is wrong with systemd | 19:53 |
| openstackgerrit | Merged openstack-infra/elastic-recheck master: Add query for neutron dhcpv6 bug 1655567 https://review.openstack.org/470416 | 19:53 |
| openstack | bug 1655567 in neutron "DHCPv6 failures with "address already allocated in subnet" error" [High,In progress] https://launchpad.net/bugs/1655567 - Assigned to Miguel Lavalle (minsel) | 19:54 |
| *** zz_dimtruck is now known as dimtruck | 19:54 | |
| dirk | clarkb: once more +2 please, had to fix a syntax error.. | 19:55 |
| *** eharney has quit IRC | 19:56 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/devstack-gate master: List swift and keystone in feature.yaml https://review.openstack.org/470383 | 19:58 |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config master: Only run swift and keystone in swift func jobs https://review.openstack.org/470421 | 19:58 |
| pabelanger | sorry, audit email for zuul user was me | 19:58 |
| openstackgerrit | Matthew Thode proposed openstack-infra/project-config master: Create the requirements stable maintence team https://review.openstack.org/470419 | 19:58 |
| clarkb | dirk: done | 19:59 |
| pabelanger | ah, likely firewall rules need to be reloaded on nodepool.o.o since we changed IPs | 20:00 |
| pabelanger | think we are not connecting to zookeeper | 20:00 |
| fungi | if the rules reference server names in dns rather than specific ip addresses, and you updated the address records in dns after the firewall rule additions were applied, then you'll need to reload them yes | 20:01 |
| pabelanger | that is better | 20:03 |
| pabelanger | zuulv3.o.o now listening to events from gerrit | 20:03 |
| pabelanger | jeblair: I did notice this exception on startup: http://paste.openstack.org/show/611340/ | 20:03 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Make finger port configurable https://review.openstack.org/470243 | 20:03 |
| clarkb | doesn't zookeeper run on nodepool.o.o? | 20:04 |
| pabelanger | it does | 20:04 |
| clarkb | oh you mean zuulv3 ip changed | 20:04 |
| pabelanger | I had to reload its iptables | 20:04 |
| clarkb | I read it as nodepool ip changed | 20:04 |
| *** harlowja has joined #openstack-infra | 20:04 | |
| mordred | pabelanger: that paste seems un-nice | 20:04 |
| pabelanger | yes, first server was trusty | 20:04 |
| clarkb | roger | 20:04 |
| pabelanger | had to recreate as xenial | 20:04 |
| pabelanger | but, we are running under python3 | 20:04 |
| pabelanger | systemd script will likely need some updates | 20:05 |
| pabelanger | we might just want to move directly to systemd, vs xenial convering init sript | 20:05 |
| openstackgerrit | Merged openstack-infra/puppet-openstackid master: Updated session cookie params https://review.openstack.org/470407 | 20:05 |
| *** mrunge has quit IRC | 20:06 | |
| *** dprince has quit IRC | 20:06 | |
| pabelanger | jeblair: clarkb: fungi: do you mind reviewing https://review.openstack.org/#/c/469995/ and https://review.openstack.org/#/c/469994/ logic needed to bring online a zuul-executor | 20:06 |
| clarkb | looks like openstackclient runtimes are still killing our devstack runtimes | 20:06 |
| mordred | clarkb: yeah? | 20:06 |
| clarkb | dtroyer: mtreinish ^ do you know if the idea of putting osc in a venv to reduce the python libs it had to scan on invocation ever went anywhere? | 20:06 |
| clarkb | mordred: ya like 2-3 seoncds on each call | 20:07 |
| mordred | clarkb: jeez | 20:07 |
| mordred | pabelanger: seems like witha xenial target just doing systemd would be the right choice | 20:07 |
| mordred | clarkb: ^^ ? | 20:07 |
| *** mrunge has joined #openstack-infra | 20:07 | |
| clarkb | ++ | 20:08 |
| clarkb | systemd is a pretty nice init system | 20:08 |
| clarkb | but you have to embrace it and deal with th warts to get there | 20:08 |
| * mordred examines clarkb for signs of compromise | 20:08 | |
| clarkb | :) | 20:08 |
| mordred | clarkb: how long have you been a vegetarian? | 20:09 |
| clarkb | I mean I don't like that it does ntp and still on the fence about journald, and nfs mounts etc | 20:09 |
| clarkb | but as a run these things on boot system its kinda nice | 20:09 |
| clarkb | mordred: I had sausages for lunch yesterday and dinner, so uh since dinner | 20:09 |
| clarkb | oh and dns, why does systemd need a dns server that resolves over dbus? | 20:10 |
| *** marst has quit IRC | 20:10 | |
| pabelanger | ya, I've been playing with systemd service config files too recently. | 20:10 |
| mordred | clarkb: why does systemd have a dns server that resolves incorrectly over dbus | 20:10 |
| *** marst has joined #openstack-infra | 20:10 | |
| pabelanger | you'd creata /etc/systemd/system/zuul-executor.service.d/override.conf to customize the systemd script for exmaple | 20:10 |
| * dtroyer sees the solution to the world's problemsL a systemd plugin that runs a pre-loaded OSC | 20:11 | |
| clarkb | infra-root as dirk points out we have a handful of d-g changes that could be reviweed. Some likely want sdague input as it touches journald logging but otherwise reviews would be appreciated | 20:11 |
| clarkb | dtroyer: I like it | 20:11 |
| mordred | clarkb: don't know if you've seen, but there's a bunch of threads about the bugs in systemd dns resolution, dns poisining attacks, and upstream refusing to fix | 20:11 |
| clarkb | dtroyer: we can send commands to the interactive shell in osc over dbus | 20:11 |
| mordred | dtroyer: ++ | 20:11 |
| clarkb | mordred: I hadn't I think dns over dbus is a generally bad idea regardless | 20:11 |
| mordred | clarkb: + | 20:11 |
| *** eharney has joined #openstack-infra | 20:11 | |
| dtroyer | seriously, we can do it via subprocesses today if you can correctly guess where output ends | 20:12 |
| mordred | pabelanger: do we care about the rspec failures related to nodepool on the first patch? | 20:12 |
| clarkb | dtroyer: if running as a systemd unit output will end up in journald | 20:12 |
| openstackgerrit | Hongbin Lu proposed openstack-infra/project-config master: [Zun] Remove nova-docker non-voting job https://review.openstack.org/470424 | 20:12 |
| *** rcernin has joined #openstack-infra | 20:12 | |
| mordred | pabelanger: also, yah - I think we should unpin | 20:13 |
| clarkb | dtroyer: biggest issue will likely be distinguishing between old and current output ( doable just neesd to be sorted out) | 20:13 |
| clarkb | dtroyer: might be easier to have it overwrite a file on each command? | 20:13 |
| clarkb | dtroyer: I think the runtime gets a lot better just be isntalling osc into a venv though as that reducing the python deps that pkg_resources has to scan | 20:13 |
| pabelanger | mordred: ya, xenail rspec is currently broken. I need to port some patch from cmurphy still | 20:13 |
| dtroyer | the current approach is to use null-terminated command output | 20:14 |
| clarkb | dtroyer: oh that works too, then just get the last entry. Simple I like it | 20:14 |
| cmurphy | pabelanger: mordred yeah we're not taking very good care of them | 20:14 |
| *** stevemar__ has quit IRC | 20:14 | |
| dtroyer | re the venv bit, my speed tests were all done in a venv and load times still approached 1.5sec | 20:15 |
| dtroyer | the scanning is noticable but not the long pole, maybe in the top 5 | 20:15 |
| clarkb | dtroyer: ya since osc has a few deps of its own its not a perfect fix, but compared to all of openstack its much better | 20:15 |
| *** stevemar__ has joined #openstack-infra | 20:15 | |
| clarkb | dtroyer: the scanning was the biggest time sink when I profile it | 20:15 |
| clarkb | well and the sorting | 20:15 |
| clarkb | bceause it also sorts the giant list of things | 20:15 |
| clarkb | https://github.com/pypa/setuptools/issues/510 has like 4 implementations of hackarounds by different people | 20:17 |
| *** stevemar__ has quit IRC | 20:20 | |
| dirk | clarkb: that can be easily fixed in pbr though I think | 20:20 |
| clarkb | osc has the problem where it relies on pkg_resources for entrypoints right? | 20:20 |
| *** yamamoto has joined #openstack-infra | 20:20 | |
| clarkb | dirk: ya the issue is ^ | 20:20 |
| clarkb | dirk: so we can fix it but then none of the osc commands will work :) | 20:20 |
| dtroyer | yes… we get rid of that and the py2exe class of tools become viable | 20:21 |
| clarkb | but maybe I'm mistaken and the entrypoint loading is indepednent of the boot up scanning | 20:21 |
| dirk | clarkb: iirc its the import pbr.version | 20:21 |
| clarkb | dirk: its anything that imports pkg_resources | 20:21 |
| dirk | clarkb: which does some crazyness to figure out the version number. | 20:21 |
| dirk | clarkb: sure, but only pbr.version does that | 20:21 |
| clarkb | dirk: so while you can stop it in pbr it doesn't actually fix it | 20:21 |
| clarkb | no osc uses it as a fundamental construct for adding commands | 20:22 |
| clarkb | even the built in commands are registered that way | 20:22 |
| *** Goneri has quit IRC | 20:22 | |
| *** dhajare-brb has quit IRC | 20:22 | |
| fungi | in theory, pbr.version calls could be guarded behind a conditional and only traversed when --version is passed right? | 20:22 |
| mordred | yah | 20:22 |
| clarkb | I actually psuehd a patch to pbr related to this then sdague and dtroyer did digging and it isn't really feasible iirc | 20:23 |
| clarkb | beacuse osc itslef needs to change to see the difference | 20:23 |
| dtroyer | cliff is entirely built on entrypoints to do the command->class mapping | 20:23 |
| dirk | fungi: no, the version number is being part of the wrapper/plugin interface | 20:23 |
| dirk | fungi: so it is always determined as first thing | 20:24 |
| dirk | clarkb: interesting.. I didn't quite see that way when I looked at it | 20:24 |
| mordred | the pkg_resource sort scan should at least be cached so it's only once per process invocation, right? | 20:24 |
| dirk | and pbr.version wouldn't really need pkg_resources for determining its own version number. the consolescripts could just contain the version number | 20:24 |
| *** wolverineav has joined #openstack-infra | 20:24 | |
| clarkb | mordred: yes that is already the case | 20:24 |
| mordred | clarkb: nod | 20:25 |
| openstackgerrit | Merged openstack-infra/shade master: Remove designate client from shade's dependencies https://review.openstack.org/470366 | 20:25 |
| mordred | dirk: so - the reason pbr uses pkg_resources ot find the version number isso that we don't have to put versoins into python files - because we already put versions in git tag and then generate them at build time | 20:25 |
| dirk | mordred: sure, but even once is too much. when osc is installed in a normal environment it can add 1-2s of overhead | 20:25 |
| mordred | THAT SAID - we could generate a .py file at build time instead of the metadata file we put into egg_info | 20:25 |
| *** yamamoto has quit IRC | 20:26 | |
| clarkb | dirk: right but what we are saying is you cannot change that in pbr without also rewriting massive portions of osc and all of its plugins | 20:26 |
| clarkb | dirk: so we could make pbr better and anything not otherwise using pkg_resources would be happy | 20:26 |
| clarkb | but it won't fix this for osc | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_post_checkout https://review.openstack.org/470392 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Rename TestCloner TestExecutorRepos https://review.openstack.org/470393 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_post_checkout https://review.openstack.org/470381 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_post_and_master_checkout https://review.openstack.org/470385 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_periodic https://review.openstack.org/470053 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_git_cache_dir https://review.openstack.org/470389 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_upgrade https://review.openstack.org/470390 | 20:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove test_periodic_update https://review.openstack.org/470391 | 20:26 |
| *** Apoorva_ has quit IRC | 20:26 | |
| fungi | wall-o-zuulv3 | 20:27 |
| dirk | clarkb: understood, I didn't find that problem yet, but it would still solve a large part of the issues imho | 20:27 |
| *** Apoorva has joined #openstack-infra | 20:27 | |
| dirk | clarkb: when using any system-wide installed python*client you have 1-2s of overhead (worse when /usr is on NFS) | 20:27 |
| dirk | mordred: clarkb : https://bugs.launchpad.net/pbr/+bug/1688405 | 20:28 |
| openstack | Launchpad bug 1688405 in PBR "pbr.version slowness due to excessive setuptools/pkg_resources usage" [Undecided,New] | 20:28 |
| mordred | well - I'll work on a patch to see about pbr injecting a .py file into the project's package at build time and loading from it if it exists | 20:28 |
| dhellmann | dirk : is that to get the version? or is that the scan in osc? | 20:28 |
| dirk | mordred: why not simply have the consolescripts contain a __pbr__version global that is being returned immediately from pbr.version? | 20:28 |
| fungi | mordred: that's something i'd love to see even if it's not directly solving this particular problem | 20:28 |
| clarkb | dhellmann: well its both, pkg_resources only runs it once | 20:28 |
| dirk | dhellmann: its just for finding it's own version number | 20:28 |
| mordred | dirk: because things use pbr as libraries | 20:29 |
| clarkb | dhellmann: so if you hit pbr version first it happens there, if you hit osc plugin loading it happens there | 20:29 |
| dhellmann | ah, pkg_resources loads all of the metadata? | 20:29 |
| dirk | mordred: so when the global version isn't available, fall back to pkg_resources? | 20:29 |
| mordred | dirk: so we could add the version to the console script - but if we're going to do that, we should also add a version into the library pacakge too | 20:29 |
| dirk | dhellmann: yeah, basically it stats 9000 files in a systemwide openstack installation | 20:29 |
| mordred | dirk: it's literally the same amount of work to do it for anything, so we might as well just do it | 20:29 |
| dirk | dhellmann: .. and opens.. | 20:30 |
| dhellmann | dirk : I expected that for the plugin scan, but not for the version lookup. I would have thought it could go directly to the file it needs to get the version info | 20:30 |
| mordred | dhellmann: pkg_resources frequently does things I dont' expect | 20:30 |
| dirk | dhellmann: unfortunately not | 20:30 |
| clarkb | dhellmann: pkg_rsources on import scans all python packages, and then sorts them based on pep440 versioning rules | 20:30 |
| dirk | because it doesn"t know it's own package name | 20:30 |
| clarkb | dhellmann: so as you add more and more python pacakges the time to start any python process that imports pkg_resources increases | 20:31 |
| dirk | dhellmann: so it reads all installed eggs in order to figure out if any of them provides something that it recognizes, and then uses that one to search for its version | 20:31 |
| clarkb | and as dirk says because its fs dependent doing that over something like nfs isslow | 20:31 |
| clarkb | https://github.com/pypa/setuptools/issues/510 is the upstream bug | 20:31 |
| dirk | I was looking at using appinfo for determining the version number, which doesn't do such crazyness | 20:32 |
| dirk | but simply reads the egg.info directly | 20:33 |
| dirk | the issue is that it seemed impossible to figure out our own package name | 20:33 |
| dirk | and appinfo requires that | 20:33 |
| mordred | yes. that's why we pass the package name to the version invocation call | 20:33 |
| mordred | dirk: __version__ = pbr.version.VersionInfo('python-novaclient').version_string() | 20:33 |
| *** claudiub has joined #openstack-infra | 20:34 | |
| mordred | dirk: but we can't use appinfo - pbr can't have dependencies on other libraries | 20:34 |
| dirk | mordred: yeah, but thats not the package name | 20:34 |
| dirk | mordred: package name is python_novaclient | 20:34 |
| dirk | that was exactly what I stumbled over | 20:35 |
| dirk | it works for pkg_resources because it finds python_novaclient because it provides python-novaclient | 20:35 |
| mordred | dirk: it's really not | 20:35 |
| mordred | thepackage name is really python-novaclient | 20:35 |
| mordred | that's whatyou pip install | 20:35 |
| dirk | ok, wrong nomenclature | 20:36 |
| openstackgerrit | Matthew Thode proposed openstack-infra/project-config master: Create the requirements stable maintence team https://review.openstack.org/470419 | 20:36 |
| fungi | note that as part of the pep-517 discussion on distutils-sig, there is talk about requiring package names to normalize - to _ similar to how wheel filenames get it enforced now | 20:36 |
| mordred | (sorry, this is one of those places where there's a ton of overlap, so we have to be super specific) | 20:36 |
| dirk | mordred: this is what I mean: | 20:36 |
| fungi | that could be a fun transition if that comes to pass | 20:36 |
| clarkb | dhellmann: its interseting to look at python -m cProfile foo.py where foo.py is just import pkg_resources | 20:36 |
| dirk | grep Name: /usr/lib/python2.7/site-packages/python_novaclient-6.0.0-py2.7.egg-info/PKG-INFO | 20:36 |
| dirk | Name: python-novaclient | 20:36 |
| mordred | yes | 20:36 |
| mordred | so - gimme a sec to make a patch to add a python file into the python package itself, and we can avoid this whole mess | 20:37 |
| fungi | the egg-info directory gets - normalized to _ right | 20:37 |
| mordred | it does | 20:37 |
| fungi | so as to avoid ambiguity since it expects - to be a separator | 20:37 |
| clarkb | anyways I bring this all up again because there is a patch to d-g to change the "devstack is running this takes 10-15 minutes" message and make it " | 20:38 |
| clarkb | "devstack is running this takes up to 30 minutes" and on quick survey its not wrong | 20:38 |
| pabelanger | jeblair: think I am going to duck out for some earily dinner. zuul-scheduler is running on zuulv3.o.o, next up is 469995 for ze01.o.o. We should be in good position early next week to start running jobs until bubblewrap | 20:39 |
| *** gardlt has quit IRC | 20:40 | |
| clarkb | I should lunch | 20:40 |
| dirk | clarkb: yeah, it's not a devstack only issue. It's also an end user issue.. | 20:40 |
| * mordred will have patch soon .. | 20:40 | |
| *** gouthamr has quit IRC | 20:41 | |
| *** wolverineav has quit IRC | 20:41 | |
| fungi | smarcet: the cookie settings change is now live on openstackid-dev... let me know if/when you want to start updating production openstackid.org again | 20:41 |
| dhellmann | clarkb : dtroyer and I have discussed a daemon mode for osc, using something other than pkg_resources to load plugins, and having cliff build a cache of the plugin info discovered via pkg_resources. All of them have their challenges. A daemon mode is easy to add, but involves a lot of devstack changes. The others break more things. | 20:41 |
| clarkb | dirk: yup | 20:41 |
| dirk | mordred: cool thanks.. don't forget the bug reference.. | 20:42 |
| smarcet | fungi: thx :) i will cheers | 20:42 |
| *** gordc has quit IRC | 20:42 | |
| clarkb | dhellmann: ya I don't think we'll solve it today, but its an issue that probably deserves attention. As dirk says this affects end users and our test suite | 20:43 |
| * fungi disappears into the wilds of his yard wielding tools of lawn destruction | 20:43 | |
| dhellmann | clarkb : yep | 20:43 |
| *** nicolasbock has quit IRC | 20:43 | |
| dirk | mordred: ideally it would avoid pkg_resources then all together btw, as that library imports a ton of stuff setting up almost 500 regular expressions which eats like 50ms for figuring out the email address (which is then ignored) | 20:44 |
| clarkb | dhellmann: we mgiht also push harder against upsteram? for example why does this have to be sorted upfront? | 20:44 |
| dhellmann | is the sort the expensive part? | 20:45 |
| dhellmann | I would have thought reading all those files would take longer | 20:45 |
| clarkb | dhellmann: reading the files is the slowest part | 20:45 |
| clarkb | but the sort is painful too | 20:45 |
| dhellmann | if we were going to build a cache file, where would we put it? and how would we update it? | 20:45 |
| dirk | dhellmann: cpu time is eaten by pulling in a ton of libraries for parsing pkg-infos compiling over 500 regexes | 20:46 |
| clarkb | 18 0.000 0.000 0.094 0.005 __init__.py:1977(_by_version_descending) out of 1 0.001 0.001 0.232 0.232 foo.py:1(<module>) locally | 20:46 |
| dhellmann | I guess we can look at directories on PYTHON_PATH to see if they are newer than the cache file | 20:46 |
| clarkb | .094 ms vs .0232 total just to import pkg_resources (and I have a fairly quick ssd) | 20:46 |
| dhellmann | dirk : yeah, I'm trying to figure out how to build something into stevedore (a wrapper for pkg_resources) to manage a cache file | 20:46 |
| clarkb | the rest of the time is largely in the raed offof disk | 20:46 |
| dhellmann | take all of the information in the pkg_resources working_set and save it, so we can reload it from one file instead of many | 20:47 |
| dirk | dhellmann: still pkg_resources would be initialized at some point, maybe unnecessarily? | 20:48 |
| dhellmann | the idea is the first time you run "openstack $anything" it would build the cache file, and from there on out it would read that and not use pkg_resources for finding or loading plugins | 20:48 |
| dirk | Ah, ok | 20:49 |
| dhellmann | so the next time you run "openstack $somethingelse" it's faster | 20:49 |
| dhellmann | but (a) where do we put the cache and (b) how do we ensure it's updated correctly and (c) who is going to write this because ENOTIME | 20:49 |
| dirk | dhellmann: might not be easy to get right when it's used from services (that are running as different users)? | 20:50 |
| dhellmann | yep, the cache file would have to be user-specific *and* import path specific | 20:50 |
| dhellmann | because virtualenv | 20:50 |
| dhellmann | well, back to less fun work | 20:51 |
| dirk | Lol | 20:51 |
| dirk | The thing missing is pip or something like that maintaining the state in a sqlite db | 20:53 |
| clarkb | yes and it would only need to be a subset of all thei nformation currently collected I think | 20:53 |
| clarkb | then load the full set of data when you need $entrypoint | 20:53 |
| dirk | So that you just do a query ob a db rather than iteration through a large directory tree | 20:53 |
| *** dimtruck is now known as zz_dimtruck | 20:54 | |
| dirk | Sounds like a stevedore dbus service.. :-) | 20:54 |
| *** trown is now known as trown|outtypewww | 20:54 | |
| *** gardlt has joined #openstack-infra | 20:54 | |
| *** zz_dimtruck is now known as dimtruck | 20:55 | |
| dhellmann | some of the complexity of pkg_resources is because setuptools supports multiple versions of packages. I wonder if we could build a faster version that was compatible except for that? | 20:56 |
| *** salv-orlando has joined #openstack-infra | 20:58 | |
| *** gardlt has quit IRC | 20:59 | |
| *** gouthamr has joined #openstack-infra | 21:00 | |
| *** smarcet has quit IRC | 21:03 | |
| openstackgerrit | Felipe Monteiro proposed openstack-infra/project-config master: Use keystone uuid tokens to optimize test run time https://review.openstack.org/470399 | 21:04 |
| *** kaisers1 has quit IRC | 21:07 | |
| *** mat128 has quit IRC | 21:10 | |
| *** matrohon has joined #openstack-infra | 21:10 | |
| *** thorst has quit IRC | 21:11 | |
| *** thorst has joined #openstack-infra | 21:15 | |
| *** ldnunes has quit IRC | 21:15 | |
| *** eharney has quit IRC | 21:15 | |
| *** Goneri has joined #openstack-infra | 21:18 | |
| *** jascott1 has quit IRC | 21:18 | |
| *** jascott1 has joined #openstack-infra | 21:19 | |
| *** thorst has quit IRC | 21:19 | |
| *** jascott1 has quit IRC | 21:20 | |
| *** jascott1 has joined #openstack-infra | 21:20 | |
| *** felipemonteiro__ has quit IRC | 21:21 | |
| *** yamamoto has joined #openstack-infra | 21:22 | |
| *** lbragstad has quit IRC | 21:22 | |
| *** kgiusti has quit IRC | 21:23 | |
| *** jascott1 has quit IRC | 21:25 | |
| *** yamamoto has quit IRC | 21:27 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Reject some untrusted config objects https://review.openstack.org/470442 | 21:31 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Reject some untrusted config objects https://review.openstack.org/470442 | 21:34 |
| *** hashar has quit IRC | 21:35 | |
| openstackgerrit | Monty Taylor proposed openstack-dev/pbr master: Write version info into and read from a file in the package https://review.openstack.org/470444 | 21:38 |
| mordred | dirk, clarkb, dhellmann: ^^ | 21:38 |
| *** thorst has joined #openstack-infra | 21:39 | |
| mordred | dirk, clarkb, dhellmann: also, honestly, writing a package that installs a global setuptools hook that intercepts the pkg_resoucers inserts and also inserts them into *something-else* and then using thatsomething else to look things up should not be super hard | 21:40 |
| clarkb | mordred: ya thats what the ~4 or so hackarounds do in the upstream bug | 21:40 |
| clarkb | but those implementation all just ignore the data which we need so we can't use them off the shelf. But should be easy to make our own that tries to be smarter, making it actually smart enough likely the tricky bit | 21:41 |
| *** dimtruck is now known as zz_dimtruck | 21:41 | |
| mordred | clarkb: considering that everything in openstack installs pbr, we could have pbr install a global hook that does what we need | 21:41 |
| mordred | clarkb: in fact, it already DOES install a global hook | 21:41 |
| mordred | we just take care to only execute code in that hook when requsted to :) | 21:41 |
| mordred | so we could be slightly more aggressive and execute a little bit of code no matter if we were requested to or not | 21:42 |
| *** smarcet has joined #openstack-infra | 21:42 | |
| dhellmann | mordred : do you mean at package installation time? | 21:42 |
| mordred | dhellmann: yah | 21:42 |
| dhellmann | interesting | 21:43 |
| mordred | dhellmann: I haven't actually given it a ton of thought yet - so the hack I think we could do may not be workable or too dangerous | 21:43 |
| mordred | dhellmann: but I'll poke a bit and come back with what some options are | 21:43 |
| mordred | it's been a few years since I'd done anything evil in pbr - seems like I'm due for another bout | 21:44 |
| dhellmann | mordred : did you have a special reason for writing that file as a python module instead of a text file and using pkgutils to load it? | 21:45 |
| dhellmann | I'm not sure the search is any simpler, really | 21:45 |
| dhellmann | if self.package isn't the actual package name, that is | 21:45 |
| clarkb | mordred: I left some comments, the one about the if else is why I -1'd | 21:47 |
| *** mdrabe has quit IRC | 21:47 | |
| mordred | clarkb: kk | 21:48 |
| mordred | dhellmann: nope. could totally be a text file too | 21:48 |
| *** funzo has joined #openstack-infra | 21:48 | |
| *** spzala has quit IRC | 21:49 | |
| *** zz_dimtruck is now known as dimtruck | 21:50 | |
| *** askb has joined #openstack-infra | 21:50 | |
| mordred | dhellmann: think that would be better? | 21:50 |
| *** spzala has joined #openstack-infra | 21:50 | |
| dhellmann | mordred : maybe? | 21:52 |
| dhellmann | you still have the problem with "what is the name of the module out of which you're trying to load the data" | 21:52 |
| dhellmann | if we don't actually know the right name, the rest doesn't make much difference | 21:52 |
| *** salv-orlando has quit IRC | 21:52 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack-dev/hacking master: Updated from global requirements https://review.openstack.org/440029 | 21:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack-infra/shade master: Updated from global requirements https://review.openstack.org/470447 | 21:53 |
| *** funzo has quit IRC | 21:53 | |
| *** dhinesh has quit IRC | 21:54 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/diskimage-builder master: Updated from global requirements https://review.openstack.org/470451 | 21:54 |
| openstackgerrit | Clark Boylan proposed openstack-infra/elastic-recheck master: Query for HOST_IP not found issue https://review.openstack.org/470452 | 21:54 |
| *** matrohon has quit IRC | 21:55 | |
| *** thorst has quit IRC | 21:58 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/puppet-openstackid master: Fixed typo on session config https://review.openstack.org/470466 | 21:58 |
| *** jkilpatr has quit IRC | 21:58 | |
| *** kaisers has joined #openstack-infra | 22:04 | |
| *** rlandy has quit IRC | 22:04 | |
| *** ihrachys has quit IRC | 22:07 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid master: Fixed base url config param https://review.openstack.org/470559 | 22:08 |
| *** rfolco has quit IRC | 22:09 | |
| *** smarcet has left #openstack-infra | 22:14 | |
| *** arahal__ has quit IRC | 22:15 | |
| *** jkilpatr has joined #openstack-infra | 22:16 | |
| *** dimtruck is now known as zz_dimtruck | 22:16 | |
| *** deafult has quit IRC | 22:18 | |
| *** lbragstad has joined #openstack-infra | 22:18 | |
| *** deafult has joined #openstack-infra | 22:19 | |
| openstackgerrit | K Jonathan Harker proposed openstack-infra/zuul feature/zuulv3: Add github reporter status-url config option https://review.openstack.org/449794 | 22:20 |
| *** zz_dimtruck is now known as dimtruck | 22:22 | |
| *** bobh has quit IRC | 22:22 | |
| *** yamamoto has joined #openstack-infra | 22:24 | |
| *** Swami has quit IRC | 22:24 | |
| clarkb | pabelanger: just noticing pabelanger-test is still in sto2, I can delete it for you if you like? | 22:25 |
| *** watersoul has quit IRC | 22:28 | |
| *** watersoul has joined #openstack-infra | 22:29 | |
| *** yamamoto has quit IRC | 22:29 | |
| mordred | dhellmann: indeed. however - I have discovered that doing this with a non-python file is MUCH harder, so I'm gonna go back to that route | 22:30 |
| *** kzaitsev1pi has joined #openstack-infra | 22:31 | |
| openstackgerrit | Merged openstack-infra/openstackid master: Fixed base url config param https://review.openstack.org/470559 | 22:32 |
| *** gardlt has joined #openstack-infra | 22:32 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Mutex repo updates and merge ops in executor https://review.openstack.org/470571 | 22:34 |
| *** cinerama` has joined #openstack-infra | 22:34 | |
| *** andreaf has quit IRC | 22:35 | |
| dhellmann | mordred : yeah, sorry, I didn't mean to throw you a curve ball | 22:35 |
| openstackgerrit | Monty Taylor proposed openstack-dev/pbr master: Write version info into and read from a file in the package https://review.openstack.org/470444 | 22:35 |
| mordred | dhellmann: no - it was a good curve ball - I liked it | 22:35 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Mutex repo updates and merge ops in executor https://review.openstack.org/470571 | 22:36 |
| mordred | dhellmann: it just turns out there is some serious black magic involved in getting the file actually included in the install through all of the install forms | 22:36 |
| dhellmann | I can only imagine | 22:36 |
| openstackgerrit | Paul Belanger proposed openstack-infra/devstack-gate master: Install stackviz from cached tarball https://review.openstack.org/457074 | 22:37 |
| pabelanger | clarkb: ya, I'll try to do that later this evening or on the weekend | 22:38 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Mutex repo updates and merge ops in executor https://review.openstack.org/470571 | 22:38 |
| clarkb | pabelanger: I can do it if you want | 22:38 |
| pabelanger | clarkb: sure if you don't mind | 22:38 |
| *** rcernin has quit IRC | 22:39 | |
| *** cinerama has quit IRC | 22:40 | |
| *** kzaitsev_pi has quit IRC | 22:40 | |
| *** deafult has quit IRC | 22:40 | |
| *** andreaf has joined #openstack-infra | 22:40 | |
| *** hwoarang has quit IRC | 22:40 | |
| *** gouthamr has quit IRC | 22:40 | |
| *** cody-somerville has quit IRC | 22:40 | |
| *** melwitt_ has joined #openstack-infra | 22:40 | |
| *** deafult has joined #openstack-infra | 22:40 | |
| *** hwoarang has joined #openstack-infra | 22:40 | |
| *** melwitt has quit IRC | 22:40 | |
| *** cody-somerville has joined #openstack-infra | 22:40 | |
| *** cody-somerville has joined #openstack-infra | 22:40 | |
| *** spzala has quit IRC | 22:41 | |
| *** spzala has joined #openstack-infra | 22:41 | |
| *** gouthamr has joined #openstack-infra | 22:41 | |
| *** mtreinish has quit IRC | 22:42 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config master: Revert "Reduce citycloud-sto2 to 40 max servers" https://review.openstack.org/470574 | 22:42 |
| clarkb | done and with that node cleaned up we can do ^ | 22:42 |
| clarkb | pabelanger: mordred jeblair fungi ^ revieing that before weekend would be great just to get it out of the way | 22:42 |
| *** sfinucan has quit IRC | 22:43 | |
| *** rbrndt has quit IRC | 22:45 | |
| *** spzala has quit IRC | 22:46 | |
| pabelanger | +3 | 22:46 |
| *** sfinucan has joined #openstack-infra | 22:46 | |
| *** xarses_ has quit IRC | 22:47 | |
| *** mtreinish has joined #openstack-infra | 22:47 | |
| *** Goneri has quit IRC | 22:47 | |
| *** jesusaur has quit IRC | 22:48 | |
| *** thorst has joined #openstack-infra | 22:49 | |
| *** thorst has quit IRC | 22:49 | |
| *** melwitt_ is now known as melwitt | 22:49 | |
| *** spzala has joined #openstack-infra | 22:50 | |
| *** MasterOfBugs has quit IRC | 22:50 | |
| *** marst_ has joined #openstack-infra | 22:52 | |
| *** jesusaur has joined #openstack-infra | 22:54 | |
| *** spzala has quit IRC | 22:54 | |
| *** thorst has joined #openstack-infra | 22:55 | |
| *** marst has quit IRC | 22:56 | |
| *** spzala has joined #openstack-infra | 22:56 | |
| openstackgerrit | Merged openstack-infra/project-config master: Revert "Reduce citycloud-sto2 to 40 max servers" https://review.openstack.org/470574 | 22:58 |
| *** esberglu has quit IRC | 22:59 | |
| *** ianychoi has quit IRC | 22:59 | |
| *** esberglu has joined #openstack-infra | 23:00 | |
| *** spzala has quit IRC | 23:00 | |
| *** marst_ has quit IRC | 23:00 | |
| *** ianychoi has joined #openstack-infra | 23:01 | |
| *** stevemar__ has joined #openstack-infra | 23:02 | |
| *** rkukura has joined #openstack-infra | 23:03 | |
| *** rkukura has quit IRC | 23:04 | |
| *** esberglu has quit IRC | 23:04 | |
| *** leakypipes has quit IRC | 23:10 | |
| *** rkukura has joined #openstack-infra | 23:11 | |
| *** jesusaur has quit IRC | 23:13 | |
| *** thorst has quit IRC | 23:14 | |
| *** xyang1 has quit IRC | 23:14 | |
| *** stevemar__ has quit IRC | 23:15 | |
| *** jesusaur has joined #openstack-infra | 23:16 | |
| *** bobh has joined #openstack-infra | 23:16 | |
| *** b00tcat_ has quit IRC | 23:17 | |
| *** dimtruck is now known as zz_dimtruck | 23:23 | |
| *** rhallisey has quit IRC | 23:25 | |
| *** yamamoto has joined #openstack-infra | 23:26 | |
| *** yamamoto has quit IRC | 23:31 | |
| *** rhallisey has joined #openstack-infra | 23:31 | |
| *** jascott1 has joined #openstack-infra | 23:43 | |
| *** rkukura has quit IRC | 23:44 | |
| *** Goneri has joined #openstack-infra | 23:45 | |
| *** xarses has quit IRC | 23:47 | |
| openstackgerrit | Monty Taylor proposed openstack-dev/pbr master: Write version info into and read from a file in the package https://review.openstack.org/470444 | 23:47 |
| *** funzo has joined #openstack-infra | 23:49 | |
| *** claudiub has quit IRC | 23:49 | |
| *** funzo has quit IRC | 23:54 | |
| mordred | clarkb, dhellmann ok. that approach has a flaw. I'll look in to new approaches next week | 23:57 |
| *** Goneri has quit IRC | 23:57 | |
| *** bobh has quit IRC | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!