| *** wshao has joined #openstack-relmgr-office | 00:05 | |
| *** dstanek has joined #openstack-relmgr-office | 00:05 | |
| *** dansmith has joined #openstack-relmgr-office | 00:07 | |
| *** gpoo has joined #openstack-relmgr-office | 00:07 | |
| *** jhesketh has joined #openstack-relmgr-office | 00:07 | |
| *** ttx has joined #openstack-relmgr-office | 00:07 | |
| *** sdague has joined #openstack-relmgr-office | 00:07 | |
| *** thingee has joined #openstack-relmgr-office | 00:07 | |
| *** dougshelley66 has joined #openstack-relmgr-office | 00:07 | |
| *** redrobot has joined #openstack-relmgr-office | 00:07 | |
| *** jgriffith has joined #openstack-relmgr-office | 00:07 | |
| *** mriedem has joined #openstack-relmgr-office | 00:07 | |
| *** mtreinish has joined #openstack-relmgr-office | 00:07 | |
| *** HenryG has joined #openstack-relmgr-office | 00:07 | |
| *** cp16net has joined #openstack-relmgr-office | 00:07 | |
| *** Kiall has joined #openstack-relmgr-office | 00:08 | |
| *** Einherjer has joined #openstack-relmgr-office | 00:08 | |
| *** dtantsur|afk has joined #openstack-relmgr-office | 00:08 | |
| *** sileht has joined #openstack-relmgr-office | 00:08 | |
| *** fungi has joined #openstack-relmgr-office | 00:08 | |
| *** flaper87 has joined #openstack-relmgr-office | 00:08 | |
| *** kragniz has joined #openstack-relmgr-office | 00:08 | |
| *** bnemec has joined #openstack-relmgr-office | 00:08 | |
| *** SergeyLukjanov has joined #openstack-relmgr-office | 00:08 | |
| *** freerunner has joined #openstack-relmgr-office | 00:08 | |
| *** peterstac has joined #openstack-relmgr-office | 00:08 | |
| *** SlickNik has joined #openstack-relmgr-office | 00:08 | |
| *** stevebaker has joined #openstack-relmgr-office | 00:08 | |
| *** jd__ has joined #openstack-relmgr-office | 00:08 | |
| *** patchbot has joined #openstack-relmgr-office | 00:08 | |
| *** johnthetubaguy has joined #openstack-relmgr-office | 00:10 | |
| *** david-lyle has joined #openstack-relmgr-office | 00:11 | |
| *** jraim has joined #openstack-relmgr-office | 00:11 | |
| *** morgan has joined #openstack-relmgr-office | 00:11 | |
| *** dougwig has joined #openstack-relmgr-office | 00:11 | |
| *** mikal has joined #openstack-relmgr-office | 00:12 | |
| *** smcginnis has joined #openstack-relmgr-office | 00:12 | |
| *** openstackgerrit has joined #openstack-relmgr-office | 00:12 | |
| *** bdemers has joined #openstack-relmgr-office | 00:12 | |
| *** jokke_ has joined #openstack-relmgr-office | 00:12 | |
| *** dhellmann has joined #openstack-relmgr-office | 00:12 | |
| *** notmyname has joined #openstack-relmgr-office | 00:12 | |
| *** harlowja has joined #openstack-relmgr-office | 00:12 | |
| *** vipul has joined #openstack-relmgr-office | 00:12 | |
| *** bknudson has joined #openstack-relmgr-office | 00:12 | |
| *** devananda has joined #openstack-relmgr-office | 00:12 | |
| *** anteaya has joined #openstack-relmgr-office | 00:12 | |
| *** harlowja has quit IRC | 00:13 | |
| *** harlowja has joined #openstack-relmgr-office | 00:13 | |
| *** armax has joined #openstack-relmgr-office | 00:13 | |
| *** doug-fish has joined #openstack-relmgr-office | 00:13 | |
| *** Daviey has joined #openstack-relmgr-office | 00:13 | |
| *** mugsie has joined #openstack-relmgr-office | 00:13 | |
| *** david-lyle has quit IRC | 00:14 | |
| *** nikhil_k has joined #openstack-relmgr-office | 00:17 | |
| *** dims_ has joined #openstack-relmgr-office | 00:17 | |
| *** krotscheck has joined #openstack-relmgr-office | 00:17 | |
| *** tonyb has joined #openstack-relmgr-office | 00:17 | |
| *** sigmavirus24_awa has joined #openstack-relmgr-office | 00:17 | |
| *** jroll has joined #openstack-relmgr-office | 00:17 | |
| *** mattoliverau has joined #openstack-relmgr-office | 00:17 | |
| *** isviridov_away has joined #openstack-relmgr-office | 00:17 | |
| *** dims_ has quit IRC | 00:20 | |
| *** wshao has quit IRC | 00:23 | |
| *** harlowja has quit IRC | 00:24 | |
| *** dims_ has joined #openstack-relmgr-office | 00:26 | |
| *** dims_ has quit IRC | 00:27 | |
| *** dims_ has joined #openstack-relmgr-office | 00:27 | |
| *** ChanServ has joined #openstack-relmgr-office | 00:30 | |
| *** cameron.freenode.net sets mode: +o ChanServ | 00:30 | |
| *** harlowja has joined #openstack-relmgr-office | 00:31 | |
| *** stevemar_ has joined #openstack-relmgr-office | 00:31 | |
| *** dougwig has quit IRC | 01:04 | |
| *** armax has quit IRC | 01:19 | |
| *** david-lyle has joined #openstack-relmgr-office | 01:21 | |
| *** dougwig has joined #openstack-relmgr-office | 01:40 | |
| *** spzala has joined #openstack-relmgr-office | 01:45 | |
| *** spzala has quit IRC | 02:02 | |
| *** dims_ has quit IRC | 02:14 | |
| *** dims_ has joined #openstack-relmgr-office | 02:14 | |
| *** dims_ has quit IRC | 02:49 | |
| *** dims_ has joined #openstack-relmgr-office | 02:49 | |
| *** lifeless has joined #openstack-relmgr-office | 02:51 | |
| *** dims_ has quit IRC | 02:51 | |
| *** dims_ has joined #openstack-relmgr-office | 02:51 | |
| *** armax has joined #openstack-relmgr-office | 03:28 | |
| *** dims_ has quit IRC | 04:29 | |
| *** gpoo has quit IRC | 05:30 | |
| *** dims_ has joined #openstack-relmgr-office | 06:26 | |
| *** dims_ has quit IRC | 06:28 | |
| *** stevemar_ has quit IRC | 06:35 | |
| *** stevemar_ has joined #openstack-relmgr-office | 06:36 | |
| *** stevemar_ has quit IRC | 06:38 | |
| *** armax has quit IRC | 06:53 | |
| *** dims_ has joined #openstack-relmgr-office | 07:28 | |
| *** dims_ has quit IRC | 07:33 | |
| *** dtantsur|afk is now known as dtantsur | 07:48 | |
| ttx | dhellmann: oh and while I think about it Task 4 would be to enable the periodic jobs on stable/liberty, once we get all the branches ready and the ACLs are restored | 08:03 |
|---|---|---|
| *** dims_ has joined #openstack-relmgr-office | 08:29 | |
| *** dims_ has quit IRC | 08:35 | |
| openstackgerrit | Julia Varlamova proposed openstack/releases: Stable/kilo Keystoneclient release https://review.openstack.org/235794 | 09:01 |
| openstackgerrit | Julia Varlamova proposed openstack/releases: Stable/kilo Keystoneclient release https://review.openstack.org/235794 | 09:13 |
| *** dtantsur is now known as dtantsur|brb | 09:17 | |
| *** dims_ has joined #openstack-relmgr-office | 09:31 | |
| SergeyLukjanov | ttx, I think we can merge it https://review.openstack.org/#/c/235133/ | 09:35 |
| *** dims_ has quit IRC | 09:37 | |
| *** dims_ has joined #openstack-relmgr-office | 09:48 | |
| *** dtantsur|brb is now known as dtantsur | 10:11 | |
| sdague | ttx: is there a reason not to run periodic stable jobs way earlier in the process? | 11:06 |
| ttx | sdague: not that I can think of | 11:24 |
| ttx | I think we do it after release so that we have one less thing to do pre-release (and the branch is exercised enough prerelease | 11:24 |
| ttx | ) | 11:24 |
| sdague | sure, that's fair | 11:24 |
| ttx | sdague: btw I'm leaving today for one week vacation, so don't count on me too much. dhellmann takes release-related matters and thingee takes design summit matters | 11:25 |
| sdague | ttx: no worries | 11:25 |
| sdague | flaper87 has all the creds needed for the cross project sessions, right? | 11:25 |
| ttx | sdague: yes and thingee too | 11:26 |
| sdague | I might want to update a bit of wording on a few of those | 11:26 |
| sdague | ok, cool | 11:26 |
| sdague | ttx: vacation in jp? or just time off locally | 11:26 |
| ttx | yeah, I wanted to update the one on documenting the openstack way so that we can include further discussing how to make following dev news more convenient | 11:26 |
| ttx | leaving for Japan on Sunday | 11:26 |
| * ttx adjusts wording now | 11:27 | |
| ttx | sdague: does "Documenting and keeping up with OpenStack development" sound ok to expand the scope of the "Documenting the OpenStack way" session ? | 11:28 |
| sdague | sure | 11:30 |
| ttx | ok pushed | 11:31 |
| ttx | I fixed a number of etherpad links too | 11:38 |
| ttx | I'll remove my -2s from reviews on stable/liberty so that nothing gets blocked | 11:44 |
| *** stevemar_ has joined #openstack-relmgr-office | 11:53 | |
| ttx | ok done | 11:55 |
| *** stevemar_ has quit IRC | 11:55 | |
| *** spzala has joined #openstack-relmgr-office | 11:56 | |
| sdague | great, thanks | 11:56 |
| *** gordc has joined #openstack-relmgr-office | 12:28 | |
| gordc | ttx: is there something required from ceilometer to clean up stable/liberty branch? seems like it's blocking it becuase of semvar issue | 12:42 |
| gordc | http://logs.openstack.org/36/235536/2/check/gate-requirements-integration-dsvm/7b0e14c/console.html#_2015-10-16_06_16_20_387 | 12:42 |
| ttx | gordc: sorry I haven't been following that, maybe dhellmann knows | 12:42 |
| gordc | ttx: kk. i'll wait for dhellmann | 12:43 |
| *** bswartz1 has joined #openstack-relmgr-office | 13:03 | |
| *** bswartz1 is now known as bswartz | 13:04 | |
| *** mestery has joined #openstack-relmgr-office | 13:07 | |
| sdague | gordc: that's the drop of your setup.cfg line | 13:13 |
| sdague | gordc: this is the next patch you have to land https://review.openstack.org/#/c/235815/ | 13:14 |
| sdague | which looks like you are going to have to squash requirements updates into that | 13:14 |
| sdague | because those didn't get landed | 13:15 |
| sdague | http://logs.openstack.org/15/235815/2/check/gate-ceilometer-requirements/4fd1bd4/console.html#_2015-10-16_10_22_52_431 | 13:15 |
| gordc | sdague: we can't land that patch because it's not in global-requirements. | 13:16 |
| dhellmann | gordc: see my email thread about this http://lists.openstack.org/pipermail/openstack-dev/2015-October/077156.html | 13:16 |
| gordc | our gate is also blocked because oslo.messaging 2.6.x are broken. | 13:17 |
| sdague | gordc: right, you're going to have to squash that patch with the requirements update which is missed | 13:17 |
| sdague | gordc: ok, I guess tonyb updated the patch incorrectly with requirements bits that weren't in gr | 13:19 |
| sdague | sorry, took me a second to wrap that all | 13:20 |
| dhellmann | nikhil_k: we need travis because he's the only one with permission to add anyone to the group | 13:20 |
| sdague | is the o.m block on stable/liberty proposed | 13:20 |
| gordc | sdague: right. (i was googling what i needed to squash) | 13:20 |
| dhellmann | ttx: ack on task 4 | 13:20 |
| jokke_ | Anyone planning to change the stable/liberty acls to stable-maint anytime soon? | 13:20 |
| sdague | gordc: can you squash these 2 down - https://review.openstack.org/#/c/235536 | 13:21 |
| sdague | doing that as 2 patches is just more races that can hurt us | 13:21 |
| sdague | oh, neat, those are blocked on ceilometer | 13:21 |
| gordc | sdague: yeah... it's a circular block. | 13:21 |
| nikhil_k | dhellmann: ack. if you need me to update anything I am available. | 13:22 |
| gordc | i believe the oslo.messaging bug is actually fix so maybe it's easiest to just release oslo.messaging 2.6.3 | 13:22 |
| sdague | so... the thing that went wrong is that ceilometer cut a release in this state | 13:22 |
| ttx | dhellmann: about to leave for Paris now, anything urgent ? | 13:22 |
| dhellmann | I think the o.m release came after that, didn't sdague ? | 13:22 |
| sdague | which I guess is a non tested construct | 13:22 |
| dhellmann | ttx: no, I'm just coming online but I think I can catch up from sdague | 13:23 |
| gordc | it's really because oslo.messaging is uncapped so the last few relesaes broke it | 13:23 |
| dhellmann | sdague: are constraints not working in the stable branch? | 13:23 |
| gordc | https://bugs.launchpad.net/ceilometer/+bug/1505730 | 13:23 |
| openstack | Launchpad bug 1505730 in Ceilometer "oslo.messaging 2.6.1 breaks ceilometer unit tests" [Undecided,New] | 13:23 |
| sdague | dhellmann: it's a unit test failure | 13:23 |
| dhellmann | let me start over | 13:24 |
| dhellmann | can we land the patch to cap or exclude oslo.messaging in stable/liberty? | 13:24 |
| sdague | no | 13:24 |
| dhellmann | in the global requirements | 13:24 |
| dhellmann | ok, what's blocking that? | 13:24 |
| sdague | ceilometer | 13:24 |
| dhellmann | how? | 13:24 |
| dhellmann | what's the failure? | 13:24 |
| sdague | not having landed the setup.cfg change | 13:24 |
| sdague | http://logs.openstack.org/36/235536/2/check/gate-requirements-integration-dsvm/7b0e14c/console.html#_2015-10-16_06_16_20_387 | 13:24 |
| gordc | dhellmann: semvar | 13:24 |
| dhellmann | so we have a patch in ceilometer after the stable branch was cut and that patch did not remove the version from setup.cfg? | 13:25 |
| sdague | we can no longer land any stable/liberty requirements until ceilometer (and possibly other projects) land those setup.cfg changes | 13:25 |
| gordc | i think if we release oslo.messaging it should clean up ceilometer gate (assuming the bug is actually fixed) | 13:25 |
| sdague | dhellmann: that patch is out there | 13:25 |
| sdague | it can't pass tests | 13:25 |
| sdague | because the oslo.messaging that gets installed breaks unit tests | 13:26 |
| dhellmann | the log you just pasted, what is that testing? | 13:26 |
| sdague | global requirements | 13:26 |
| sdague | adding the != 2.6.1 o.m bit | 13:26 |
| sdague | https://review.openstack.org/#/c/235536/ - requirements patch blocked on ceilometer | 13:27 |
| dhellmann | ok, the failure there is a semver error in ceilometer, which should not be an issue if we are testing at the branch point where the tag and version line match, so how did those get out of sync? | 13:27 |
| dhellmann | but that's a patch to a different repository | 13:28 |
| sdague | because post-versioning patch didn't land - https://review.openstack.org/#/c/235207/ | 13:28 |
| dhellmann | but there are no other patches in stable/liberty on ceilometer, so the version numbers for post and pre should match | 13:28 |
| sdague | our requirements integration ensures that the requirements for each project install | 13:28 |
| sdague | right, but https://review.openstack.org/#/c/235207/ can't land in ceilometer, because ceilometer can't pass unit tests | 13:29 |
| sdague | so https://review.openstack.org/#/c/235207/ was updated to include requirements changes which would let it pass unit tests | 13:29 |
| dhellmann | stop stop | 13:29 |
| sdague | but then it fails not having the requirements in sync | 13:29 |
| dhellmann | stop | 13:29 |
| sdague | ok | 13:29 |
| dhellmann | the patch to global requirements should be using head of openstack/ceilometer, right? | 13:29 |
| sdague | sure | 13:30 |
| *** dansmith is now known as superdan | 13:30 | |
| dhellmann | that patch has tag 5.0.0 and setup.cfg containing version=5.0.0 | 13:30 |
| dhellmann | so pbr should not be complaining about semver when it tries to install that | 13:30 |
| dhellmann | so why is pbr complaining in that job? | 13:30 |
| dhellmann | are we not testing head? | 13:30 |
| dhellmann | is something being added to the local repo? | 13:31 |
| dhellmann | does that log have a sha of what's being installed for ceilometer? | 13:31 |
| sdague | it's getting a requirements sync | 13:31 |
| sdague | so, that triggering it? | 13:31 |
| dhellmann | possibly? | 13:32 |
| sdague | because we're not installing master, we're installing master + attempted sync | 13:32 |
| dhellmann | I would have thought pbr would only look at the commits, not pending changes | 13:32 |
| dhellmann | maybe it's not doing that | 13:32 |
| dhellmann | let me try that locally | 13:32 |
| sdague | I do not know | 13:32 |
| gordc | https://review.openstack.org/#/c/235202/ this was merged... | 13:33 |
| gordc | not sure if it affects liberty branch though | 13:33 |
| dhellmann | gordc: it shouldn't | 13:34 |
| dhellmann | sdague: ok, if I sync requirements I can run "tox -e py27" in the ceilometer stable/liberty branch and it works as long as the change is not committed | 13:36 |
| dhellmann | which means the ceilometer code is installable in that state | 13:36 |
| dhellmann | however if I commit it, then I get the semver error | 13:37 |
| dhellmann | sdague: does the requirements sync commit the change locally before testing the installation? | 13:37 |
| sdague | it installs it | 13:37 |
| sdague | can you sync the requirements and then install the code? | 13:38 |
| dhellmann | yes, as long as I don't commit the change | 13:39 |
| dhellmann | where is that code in devstack? is it part of d-g? | 13:39 |
| sdague | requirements repo i believe | 13:39 |
| dhellmann | what runs the sync script, though | 13:39 |
| sdague | the requirements integration script | 13:40 |
| sdague | and yes, it appears to commit it - https://github.com/openstack/requirements/blob/eaf04e511c6d604ef84f47cc3f915ec9a4b95d9f/tools/integration.sh#L107-L114 | 13:40 |
| dhellmann | ok, so that's bad | 13:40 |
| dhellmann | I'll try a patch that removes that step and we'll see what breaks | 13:42 |
| sdague | you'll have to redo the clones later | 13:42 |
| dhellmann | yeah, I'm seeing that :-/ | 13:42 |
| dhellmann | it's not expensive to redo the update, so I'll repeat that I guess | 13:43 |
| sdague | yeh, that seems reasonable | 13:43 |
| sdague | it would be nice if there was a pbr permissive mode where it wouldn't explode under these circumstances | 13:44 |
| dhellmann | I hate this feature entirely, but haven't been able to convince lifeless to remove it | 13:45 |
| sdague | dhellmann: I am with you in hating this feature | 13:47 |
| sdague | it has only caused pain | 13:47 |
| dhellmann | sdague: this is the liberty script change: https://review.openstack.org/235924 | 13:48 |
| dhellmann | if that passes, I'll post one to master, too, and we can commit them in the right order | 13:48 |
| sdague | dhellmann: yeh, honestly, as long as they are both in flight, order doesn't hugely matter | 13:49 |
| dhellmann | wfm | 13:49 |
| * gordc doesn't understand the magic being discussed but wfm | 13:50 | |
| dhellmann | gordc: pbr tries to help you not mess up your version numbers by looking at what you're declaring the version to be and what the commit is | 13:51 |
| dhellmann | in this case, we have 5.0.0 in the commit at the branch point, and then the requirements test script is syncing a change and committing it | 13:51 |
| dhellmann | we also have a tag at the first commit of 5.0.0 | 13:51 |
| dhellmann | pbr wants you to update the version after the tag, but setup.cfg prevents that, so it's complaining about the conflict | 13:52 |
| gordc | ah... yeah i would've never found that. | 13:52 |
| dhellmann | the clue was that ceilometer was blocking the requirements update, but no changes to ceilometer should have been in that test environment | 13:53 |
| dhellmann | brb | 13:53 |
| gordc | yeah i did find that strange... thanks for the explanation | 13:53 |
| sdague | dhellmann: I think the second var is wrong in that script | 13:55 |
| sdague | off to get breakfast | 13:55 |
| SergeyLukjanov | dhellmann, should we merge https://review.openstack.org/#/c/235133/ already? | 13:56 |
| sdague | SergeyLukjanov: yes | 13:58 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:13 | |
| dhellmann | sdague: ack, thanks, fixing | 14:14 |
| dhellmann | SergeyLukjanov: yes, you need to merge that before anything else will work in stable/liberty -- you may have to squash other fixes into it | 14:15 |
| SergeyLukjanov | dhellmann, ack, thx | 14:21 |
| *** gpoo has joined #openstack-relmgr-office | 14:44 | |
| openstackgerrit | Matt Riedemann proposed openstack/releases: mitaka: release python-novaclient 2.32.0 https://review.openstack.org/232683 | 14:53 |
| mriedem | dhellmann: liberty is out the door so can we drop the -2 here https://review.openstack.org/#/c/232683/ ? | 14:54 |
| mriedem | that was a thing the nova team wanted to see released ^ | 14:54 |
| mriedem | johnthetubaguy: sdague: ^ | 14:54 |
| openstackgerrit | Matt Riedemann proposed openstack/releases: Juno: release python-glanceclient 0.14.3 https://review.openstack.org/235967 | 15:02 |
| *** bnemec has quit IRC | 15:03 | |
| *** david-ly_ has joined #openstack-relmgr-office | 15:11 | |
| *** david-lyle has quit IRC | 15:11 | |
| thingee | sdague: need some wording update for cross project sessions? | 15:12 |
| *** bnemec has joined #openstack-relmgr-office | 15:13 | |
| *** armax has joined #openstack-relmgr-office | 15:13 | |
| sdague | thingee: yes, but probably next week. brain not so good at english on friday | 15:14 |
| notmyname | thingee: FYI my plan is to do summit scheduling late today | 15:14 |
| bknudson | does openstack/requirements stable/liberty require setup.cfg change? https://review.openstack.org/#/q/project:openstack/requirements+branch:stable/liberty+status:open,n,z | 15:15 |
| bknudson | oh, never mind | 15:15 |
| bknudson | looks like https://review.openstack.org/#/c/235924/ takes care of it | 15:15 |
| sdague | bknudson: yes, all the yeses | 15:15 |
| sdague | oh, right, there is that issue as well | 15:16 |
| thingee | notmyname: ack | 15:16 |
| sdague | bknudson: we're about 10 minutes from an answer on if that fixes it for us | 15:16 |
| thingee | sdague: excellent. I'll be in UTC+08:00 next week, but email me mike@openstack.org with updates and I'll get them going. | 15:17 |
| sdague | thingee: ok, no prob | 15:18 |
| *** stevemar_ has joined #openstack-relmgr-office | 15:18 | |
| *** stevemar_ has quit IRC | 15:19 | |
| *** stevemar_ has joined #openstack-relmgr-office | 15:19 | |
| *** dtantsur is now known as dtantsur|afk | 15:24 | |
| *** armax has quit IRC | 15:40 | |
| sdague | dims_: so upper constraints failed again, I think because https://review.openstack.org/#/c/234026/ was approved and got in first | 15:43 |
| dhellmann | mriedem: I'll remove my -2, but it would be good to get the requirements issues in stable/liberty resolved before we go wild with more new releases | 15:48 |
| mriedem | dhellmann: this is not stable/liberty though, but i guess it's a crap shoot b/c u-c doesn't apply to unit test jobs | 15:50 |
| dhellmann | mriedem: right | 15:50 |
| mriedem | ugh | 15:51 |
| sdague | dhellmann: the liberty patch needs a recheck | 15:54 |
| sdague | we ended up on a corrupt node in rax | 15:54 |
| sdague | that was having filesystem errors | 15:54 |
| dhellmann | sdague: I'm looking over that log now | 15:54 |
| sdague | http://logs.openstack.org/24/235924/2/check/gate-requirements-integration-dsvm/2d1b855//logs/syslog.txt.gz#_Oct_16_14_25_02 | 15:54 |
| dhellmann | ah, yeah, rechecked | 15:55 |
| *** mriedem is now known as mriedem_lunch | 16:07 | |
| *** stevemar_ has quit IRC | 16:10 | |
| *** stevemar_ has joined #openstack-relmgr-office | 16:11 | |
| openstackgerrit | Brant Knudson proposed openstack/releases: Release keystonemiddleware 2.4.2 https://review.openstack.org/235999 | 16:12 |
| *** bnemec has quit IRC | 16:16 | |
| *** stevemar_ has quit IRC | 16:16 | |
| openstackgerrit | Brant Knudson proposed openstack/releases: keystonemiddleware 2.4.2 https://review.openstack.org/235999 | 16:22 |
| sdague | dhellmann / dims_ do either of you know how to generate the upper-constraints.txt manually? because I've failed on that front, and we're massively behind on that | 16:26 |
| sdague | https://review.openstack.org/#/c/231327/ it's a 105 line change at this point | 16:27 |
| sdague | I'll just guess that my merge resolution is right | 16:29 |
| dhellmann | sdague: I think there's a tox env to do that, but I haven't ever tried it on my own | 16:41 |
| sdague | dhellmann: so if I call tox -e generate, it complains it can find no pythons | 16:41 |
| sdague | if I call it with the params in the README, it errors on me | 16:42 |
| sdague | after 30 minutes or so | 16:42 |
| dhellmann | that's a long wait for an error message, do you have a log somewhere? | 16:43 |
| *** dims_ has quit IRC | 16:54 | |
| *** dims_ has joined #openstack-relmgr-office | 16:55 | |
| dims_ | sdague: reading backlog now, had taken the morning off :) | 16:58 |
| dims_ | sdague: so anteaya helped me last night to track down the job for the bot and the logs for the bot and all was well, the bot runs once a day only and i asked lifeless if we could run it twice a day at least. lifeless felt once a day was working ok and did not want to change that | 17:00 |
| dims_ | sdague: dhellmann: should we temporarily bump it to 3-4 times a day till the dust settles? | 17:01 |
| dhellmann | dims_, sdague: we've been putting off approving constraint updates during the release period, which may explain how far behind we are. would running the bot more often help? | 17:02 |
| dims_ | dhellmann: i'd think so, especially if a single run failed because of a transient issue | 17:03 |
| dhellmann | I'm not sure we have anyone looking at failures of the patches that bot creates, so as soon as jenkins fails it's more or less blocked | 17:03 |
| dims_ | dhellmann: only data point i have is that the run from bot day before yesterday was good. let me find the one from last night | 17:04 |
| dims_ | dhellmann: http://logs.openstack.org/periodic/propose-requirements-constraints-master/?C=M;O=D | 17:04 |
| dims_ | so last nights run succeeded | 17:05 |
| dims_ | and looks like sdague already +2A'ed it - https://review.openstack.org/#/c/231327/ | 17:06 |
| dims_ | ok i am going to watch zuul jobs for 231327 | 17:06 |
| dhellmann | dims_: I don't have a problem running the job more often, it just wasn't clear that would actually help us | 17:09 |
| *** stevemar_ has joined #openstack-relmgr-office | 17:10 | |
| dims_ | dhellmann: let's day the bot proposed a change in uc, then we approved a bunch of requirements changes (legitimate) and when the bot proposed is about to merge it finds out that the bunch of requirements changes invalidates the bot proposed review then we have to wait another day for another bot proposed revew | 17:11 |
| dims_ | and so on | 17:11 |
| dims_ | s/day/say | 17:11 |
| dhellmann | yeah, that makes sense | 17:11 |
| *** mriedem_lunch is now known as mriedem | 17:15 | |
| *** smcginnis has quit IRC | 17:30 | |
| dims_ | sdague: 231327 is now paste check into gate. | 17:44 |
| *** armax has joined #openstack-relmgr-office | 17:57 | |
| *** bnemec has joined #openstack-relmgr-office | 17:59 | |
| *** stevemar_ has quit IRC | 18:17 | |
| *** stevemar_ has joined #openstack-relmgr-office | 18:18 | |
| *** stevemar_ has quit IRC | 18:22 | |
| *** stevemar_ has joined #openstack-relmgr-office | 18:25 | |
| *** smcginnis has joined #openstack-relmgr-office | 18:38 | |
| *** dims_ is now known as dimsum__ | 18:41 | |
| lifeless | sdague: whats the error you get ? | 18:52 |
| lifeless | dhellmann: which feature do you hate? | 18:53 |
| *** smcginnis has quit IRC | 19:00 | |
| *** gordc has quit IRC | 19:02 | |
| *** david-ly_ has quit IRC | 19:04 | |
| *** david-lyle has joined #openstack-relmgr-office | 19:05 | |
| dhellmann | lifeless: http://logs.openstack.org/36/235536/2/check/gate-requirements-integration-dsvm/7b0e14c/console.html#_2015-10-16_06_16_20_387 | 19:06 |
| lifeless | dhellmann: ah. | 19:07 |
| lifeless | dhellmann: so - if we can find a sensible behaviour, I'm not attached to this one | 19:07 |
| dhellmann | lifeless: can we make that a warning instead of an exception? | 19:07 |
| lifeless | dhellmann: what version *should* pbr choose in that case? | 19:08 |
| lifeless | dhellmann: (ignoring that the warning won't be seen by anyone and other pragmatic questions) | 19:08 |
| lifeless | dhellmann: [because pip hides build output if the build succeeds] | 19:08 |
| dhellmann | well, it's clearly not 5.0, so I guess 5.0.0.devsomething | 19:08 |
| lifeless | dhellmann: thats *less* than 5, so it would be rolling the clock backwards | 19:09 |
| dhellmann | ok, 5.0.1.devsomething | 19:09 |
| lifeless | dhellmann: so the version = line in setup.cfg would be ignored? | 19:09 |
| dhellmann | yeah, if this condition is detected | 19:10 |
| lifeless | dhellmann: I think thats a backwards compat break from the definition of version that pbr's had for ages. | 19:10 |
| lifeless | dhellmann: we could add a new version-hint key or something and transition to using that | 19:11 |
| dhellmann | how old is this error handler? | 19:11 |
| dhellmann | I want us to stop using pre-versioning entirely, so if that's the answer this is just another reason to push that ahead | 19:11 |
| lifeless | dhellmann: the error came in in 0.10 IIRC | 19:11 |
| lifeless | possibly 0.11 | 19:11 |
| dhellmann | ok. we're seeing more and more cases where unintended situations get us into the error state when it really doesn't matter, so it's becoming more annoying | 19:12 |
| lifeless | so, I agree that its disruptive; I am certainly ok with a system where we hint at the version we want and then roll past it seamlessly; my concern is the existing definition of version as 'the version we're working to' which pbr has had since ~forever | 19:13 |
| lifeless | we didn't have the error in way old versions because we ran the clock backwards | 19:13 |
| lifeless | we've *never* generated a version higher than that in setup.cfg | 19:13 |
| dhellmann | so we'd use a different key in setup.cfg? | 19:13 |
| lifeless | so I think the choices are: | 19:13 |
| lifeless | - change the definition in-place hope that nothing and noone gets actually broken by it | 19:14 |
| lifeless | - add a new key with a nicer definition and let folk opt-in [and we can mass opt-in openstack projects] | 19:14 |
| lifeless | - do nothing and just stop using version= in setup.cfg's | 19:14 |
| lifeless | at least, thats all that I can think of | 19:15 |
| dhellmann | option 3 doesn't really work, since at the point we have a new stable branch we have 2 branches competing for the next post-version number. ttx's suggestion to get around that was to use pre-versioning on master just until the next release there to force the version higher than what the stable branch would compute. If we combine that with a new key that lets us roll past the version, I think that gets us to a nicer spot all around, | 19:16 |
| dhellmann | since we can mostly use post-versioning and we can use pre-versioning just during the period after a branch. | 19:16 |
| lifeless | I really don't have a good estimator for the risk of the first choice | 19:16 |
| dhellmann | I agree that option 1 is not great | 19:16 |
| lifeless | I don't see th problem with two branches competing for the next version | 19:16 |
| lifeless | Actually cutting releases that conflict would be a problem | 19:17 |
| dhellmann | patches on stable/liberty and patches on master right after stable liberty might be different but would generate the same version numbers | 19:17 |
| lifeless | sure | 19:17 |
| dhellmann | I thought for CI/CD that would introduce confusion | 19:18 |
| lifeless | there are three users of those versions that I know of | 19:18 |
| lifeless | devs, CI/CD, and redistributors | 19:18 |
| lifeless | we expect CI/CD folk to pick a branch and stick to it | 19:18 |
| lifeless | with pivots between branches (if they're not on master) around beta time | 19:18 |
| lifeless | we don't expect redistribution of non-releases on master | 19:19 |
| dhellmann | I'm describing the proposal on line 53 of https://etherpad.openstack.org/p/mitaka-release-drop-pre-versioning -- could you add comment there explaining why it's not needed? | 19:20 |
| dhellmann | because if we expect those cases not to matter, it would be good to go with the simpler solution | 19:20 |
| *** smcginnis has joined #openstack-relmgr-office | 19:28 | |
| lifeless | yeah, thanks for the c-p | 19:39 |
| lifeless | just making C breakfast | 19:39 |
| lifeless | also I've fixed up a couple of nits in the draft - e.g. postN vs devN | 19:39 |
| *** mriedem has quit IRC | 19:50 | |
| *** smcginnis has quit IRC | 19:52 | |
| dhellmann | lifeless: cool, thanks | 19:57 |
| *** smcginnis has joined #openstack-relmgr-office | 20:00 | |
| *** armax has quit IRC | 20:19 | |
| *** armax has joined #openstack-relmgr-office | 20:22 | |
| *** armax has quit IRC | 20:44 | |
| *** bswartz has quit IRC | 20:52 | |
| cp16net | dimsum__: does the recheck not work for the patch creater? https://review.openstack.org/#/c/235588/ | 21:29 |
| dimsum__ | cp16net: when you recheck you should be able to see it http://status.openstack.org/zuul/ | 21:36 |
| *** dims_ has joined #openstack-relmgr-office | 21:38 | |
| *** dimsum__ has quit IRC | 21:40 | |
| *** dims_ has quit IRC | 21:45 | |
| *** david-ly_ has joined #openstack-relmgr-office | 21:50 | |
| *** david-lyle has quit IRC | 21:50 | |
| *** armax has joined #openstack-relmgr-office | 21:53 | |
| tonyb | sdague: Yeah I deleiberately had stuff in my ceilometer fix that wasn't in g-r. I've never tried to circumvent the requiremenst stuff so I didn't really know what the gate-requirements jobs did. | 21:53 |
| tonyb | sdague: dhellmann has set us on the right path | 21:53 |
| *** stevemar_ has quit IRC | 22:01 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:01 | |
| *** stevemar_ has joined #openstack-relmgr-office | 22:02 | |
| *** stevemar_ has quit IRC | 22:06 | |
| *** stevemar_ has joined #openstack-relmgr-office | 22:07 | |
| *** dimsum__ has joined #openstack-relmgr-office | 22:08 | |
| *** stevemar_ has quit IRC | 22:12 | |
| cp16net | dimsum__: omg... i was looking for the wrong project... my bad | 22:13 |
| cp16net | i think i have a case of the fridays... | 22:13 |
| *** dimsum__ has quit IRC | 22:16 | |
| *** dimsum__ has joined #openstack-relmgr-office | 22:17 | |
| *** mestery has quit IRC | 22:24 | |
| *** bswartz has joined #openstack-relmgr-office | 22:28 | |
| openstackgerrit | Doug Hellmann proposed openstack/releases: update history import tools for liberty https://review.openstack.org/236155 | 22:34 |
| openstackgerrit | Doug Hellmann proposed openstack/releases: import liberty versions for unmanaged projects https://review.openstack.org/236156 | 22:34 |
| *** dimsum__ has quit IRC | 22:52 | |
| *** dimsum__ has joined #openstack-relmgr-office | 23:08 | |
| *** spzala has quit IRC | 23:41 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!