| openstackgerrit | Michael Johnson proposed openstack/diskimage-builder master: squash-package-install to use the correct python https://review.openstack.org/443336 | 00:59 |
|---|---|---|
| *** pmannidi has quit IRC | 01:30 | |
| openstackgerrit | Michael Johnson proposed openstack/diskimage-builder master: squash-package-install to use the correct python https://review.openstack.org/443336 | 02:39 |
| *** sudipto has joined #openstack-dib | 03:13 | |
| *** sudipto_ has joined #openstack-dib | 03:14 | |
| *** frontrunner has joined #openstack-dib | 03:15 | |
| *** sudipto has quit IRC | 03:37 | |
| *** sudipto_ has quit IRC | 03:37 | |
| *** aparnav has joined #openstack-dib | 03:43 | |
| *** frontrunner has quit IRC | 04:05 | |
| *** jamielennox is now known as jamielennox|away | 04:13 | |
| *** jamielennox|away is now known as jamielennox | 04:47 | |
| *** chhavi_ has joined #openstack-dib | 04:57 | |
| ianw | yolanda / anyone : just a heads up on what happened today | 05:13 |
| ianw | i tagged master as 2.0.0rc2 as having 2.0.0rc1 on the other branch seemed to be messing up pbr's version recognition and causing problems | 05:14 |
| ianw | there seems to have been some missed coverage in the tripleo jobs for multinode testing that hit instack | 05:14 |
| ianw | i believe that https://review.openstack.org/#/c/443367/2 is a sufficient solution | 05:14 |
| ianw | i have some ideas on how to move instack into dib itself ... but that's for another day | 05:15 |
| ianw | i think the workaround has been to pin to 1.28.0 for now | 05:15 |
| ianw | so, with all this, i will let 2.0.0rc2 sit for tomorrow and the weekend my time, and look at tagging 2.0.0 monday | 05:16 |
| ianw | that will release it to pypi and anyone not pinned | 05:16 |
| ianw | in the mean time, the dsvm jobs are fixed, so CI should be all green | 05:17 |
| *** sudipto has joined #openstack-dib | 05:25 | |
| yolanda | ianw, what it seems to fail are the tripleo jobs. They consume dib from master, not using tags | 05:40 |
| *** yolanda has quit IRC | 05:41 | |
| *** chhavi_ has quit IRC | 05:50 | |
| *** andreas-f has quit IRC | 06:45 | |
| *** andreas-f has joined #openstack-dib | 07:06 | |
| *** jamielennox is now known as jamielennox|away | 07:09 | |
| *** jamielennox|away is now known as jamielennox | 07:27 | |
| *** hashar has joined #openstack-dib | 08:03 | |
| *** chhavi_ has joined #openstack-dib | 09:50 | |
| *** hashar has quit IRC | 09:59 | |
| *** hashar has joined #openstack-dib | 10:08 | |
| openstackgerrit | Holger Smolinski proposed openstack/diskimage-builder master: Add zipl support for s390 architecture with SCSI boot https://review.openstack.org/443548 | 10:10 |
| openstackgerrit | wangjianhe proposed openstack/diskimage-builder master: download image Name suffix is error https://review.openstack.org/443559 | 10:47 |
| *** sudipto has quit IRC | 10:51 | |
| *** chhavi_ has quit IRC | 11:43 | |
| *** chhavi_ has joined #openstack-dib | 11:58 | |
| *** chhavi_ has quit IRC | 12:24 | |
| *** sudipto has joined #openstack-dib | 12:29 | |
| *** chhavi_ has joined #openstack-dib | 12:42 | |
| *** aparnav has quit IRC | 13:10 | |
| *** frontrunner has joined #openstack-dib | 13:20 | |
| *** chhavi_ has quit IRC | 13:51 | |
| *** chhavi_ has joined #openstack-dib | 13:52 | |
| *** aparnav has joined #openstack-dib | 13:56 | |
| *** strigazi has quit IRC | 14:44 | |
| *** aparnav has quit IRC | 14:47 | |
| *** strigazi has joined #openstack-dib | 15:04 | |
| *** chhavi_ has quit IRC | 15:13 | |
| *** chhavi_ has joined #openstack-dib | 16:20 | |
| *** sudipto has quit IRC | 16:44 | |
| openstackgerrit | Michael Johnson proposed openstack/diskimage-builder master: squash-package-install to use the correct python https://review.openstack.org/443336 | 16:47 |
| *** yolanda has joined #openstack-dib | 17:39 | |
| *** hashar has quit IRC | 17:55 | |
| *** trevormc has joined #openstack-dib | 18:10 | |
| *** chhavi_ has quit IRC | 18:52 | |
| *** hashar has joined #openstack-dib | 20:07 | |
| greghaynes | ianw: it seems the change to move bin/disk-image-create to an entrypoint merged in to master? | 20:22 |
| greghaynes | oh, I think I see, 2.0 got tagged on master? | 20:25 |
| greghaynes | we probably shouldnt have tagged any more 1.0's then | 20:25 |
| greghaynes | if that branch merged in | 20:25 |
| greghaynes | since techincally we didnt follow semver on them | 20:26 |
| greghaynes | and actually, we should branch v1, revert some of the backwards incompat stuff, then tag a new 1.x release on that branch | 20:27 |
| ianw | greghaynes: 2.0.0rc2 was tagged on master ... tripleo had some issues yesterday | 20:48 |
| ianw | so there isn't a 2.0.0 release yet | 20:49 |
| ianw | i'm going to wait until monday my time before unleashing it via pip like that, but tripleo should be ok | 20:49 |
| greghaynes | ianw: the issue is that the 1.x release on master after that is not following semver | 20:50 |
| greghaynes | So people got something after the 2.x rc already | 20:51 |
| greghaynes | Since 1.x is after 2.0rc | 20:51 |
| greghaynes | Oh nvm | 20:52 |
| ianw | after? there's 1.28, the merge and two fixups and then 2.0.0rc2? | 20:52 |
| greghaynes | I was looking at it out of order somehow | 20:52 |
| greghaynes | Ok whew sorry :) | 20:52 |
| ianw | if we need a new 1.x we can branch, but i'd rather do that only if we're forced | 20:52 |
| greghaynes | Oh wait, no, 2.0rc1 | 20:52 |
| greghaynes | Is before 1.28 | 20:53 |
| ianw | but there is a 2.0.0rc2? the suggestion from from release people who said to tag the final rc as the release | 20:53 |
| greghaynes | The issue is that in 1.28 I think we have some 2.x features | 20:54 |
| greghaynes | Which are not backwards compatible | 20:54 |
| greghaynes | So 1.28 is actually 2.0rc1.5 | 20:54 |
| ianw | hmm, that should not be the case at all | 20:55 |
| greghaynes | Oh so 2.0rc1 didn't have 2.0 things or? | 20:55 |
| greghaynes | That's what's confusing me, how we have any 1.x that happens after a 2.0rc in history | 20:56 |
| ianw | are you looking via gitk or something? | 20:56 |
| greghaynes | http://git.openstack.org/cgit/openstack/diskimage-builder/log/ | 20:56 |
| ianw | i think that's because that was a point where we merged master into the v2 branch | 20:57 |
| ianw | now we've merged it back the other way | 20:57 |
| greghaynes | Oh, huh | 20:57 |
| *** frontrunner has quit IRC | 20:57 | |
| ianw | but if you checkout 1.28.0 then that should just be the point before we merged | 20:58 |
| ianw | before we merged v2 into master | 20:58 |
| ianw | yeah, if i switch back on a fresh tree all the elements come back and bin/* scripts etc | 20:59 |
| greghaynes | Gotcha, sorry for the confusion :) | 20:59 |
| greghaynes | So we just need to make sure to not do any more 1.x tags on master | 20:59 |
| ianw | yeah, i think that just badly indicates the points where we updated v2 to master. there's others, but we didn't tag | 20:59 |
| ianw | yes, no more 1.x tags on master | 20:59 |
| ianw | assuming nothing else explodes today, i'm not aware of anything yet, i'll tag 2.0.0 on monday my time | 21:00 |
| greghaynes | What happened is trove exploded due to bin/disk-image-create missing | 21:00 |
| ianw | oh ... so they're using from master? | 21:00 |
| greghaynes | Which is now the case on master | 21:00 |
| greghaynes | Yep | 21:00 |
| ianw | hmmm, see i though most people would be using from requirements | 21:00 |
| ianw | anyway, a big email did go out about this requesting testing which i assume nobody read | 21:00 |
| greghaynes | Yea, that's fine though - master has to change at some point | 21:01 |
| greghaynes | That just led me to looking at log and then getting confused | 21:01 |
| ianw | but i was thinking actualy as one last 2.0.0 change i would take the details in that email and add them to a detailed relno entry | 21:01 |
| greghaynes | Sgtm | 21:02 |
| ianw | ok, loop me into any bug / change if you like. it should be very easily solved by pip installing dib which is arguably what should have been done anyway | 21:02 |
| greghaynes | Yep I think they said they'll switch to requirements | 21:02 |
| ianw | yeah, fixing this issue will help when someone bumps it too :) | 21:03 |
| greghaynes | Also, the reason I might recommend making a v1 branch even if it is no different than the latest 1.x tag is it makes it easier if someone actually wants to 1.x pin | 21:03 |
| greghaynes | They can pin to the branch now and get fixes if any happen without having to change their setup | 21:03 |
| ianw | true ... i was just assuming nobody was really going to bother with the fixes part though | 21:04 |
| ianw | https://review.openstack.org/443367 is a workaround for instack | 21:04 |
| greghaynes | Yep it's likely they'd never happen | 21:04 |
| ianw | i've sort of played around with adding a "python runner" to run arbitrary elements as part of dib to replace this | 21:04 |
| ianw | something we could actually test | 21:04 |
| greghaynes | Yes that really should live in dib | 21:05 |
| ianw | yeah, we could have basically a dib-run-parts wrapper that just sets enough environment to run the elements within your own environment | 21:05 |
| ianw | which is, afaict, what's going on here | 21:05 |
| ianw | so, anyway, not ready yet but it would be very nice to not have this hanging out there just waiting for us to break assumptions | 21:06 |
| greghaynes | ++ | 21:10 |
| greghaynes | https://review.openstack.org/#/c/443958/ is the trove change | 21:17 |
| greghaynes | ianw: im running in to another issue ATM http://paste.openstack.org/show/602122/ | 21:43 |
| greghaynes | ianw: it seems that when you install the dib module nothing is under the diskimage_builder module... | 21:43 |
| greghaynes | oh I have to import it directly | 21:44 |
| greghaynes | derp | 21:44 |
| greghaynes | this is quite the derp day for me :) | 21:44 |
| ianw | oh yeah, import diskimage_builder.paths | 22:13 |
| *** trevormc has quit IRC | 22:28 | |
| *** jamielennox is now known as jamielennox|away | 22:42 | |
| *** jamielennox|away is now known as jamielennox | 22:45 | |
| *** pmannidi has joined #openstack-dib | 23:44 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!