Friday, 2015-09-18

*** bswartz has joined #openstack-relmgr-office00:16
*** dims_ has joined #openstack-relmgr-office00:38
*** david-lyle has quit IRC00:40
*** dims_ has quit IRC00:43
*** dims_ has joined #openstack-relmgr-office00:43
*** david-lyle has joined #openstack-relmgr-office00:44
*** bdemers has quit IRC01:06
*** dims_ has quit IRC01:27
*** dims_ has joined #openstack-relmgr-office01:29
*** dims_ has quit IRC01:37
*** TravT has joined #openstack-relmgr-office01:47
*** TravT_ has joined #openstack-relmgr-office01:49
*** TravT has quit IRC01:52
*** TravT_ has quit IRC02:18
*** mriedem1 is now known as mriedem02:54
*** bdemers has joined #openstack-relmgr-office02:54
*** TravT has joined #openstack-relmgr-office02:59
*** TravT has quit IRC03:05
*** dims_ has joined #openstack-relmgr-office03:22
*** dims_ has quit IRC03:28
*** mriedem has quit IRC03:32
*** openstackgerrit has quit IRC05:31
*** openstackgerrit has joined #openstack-relmgr-office05:31
*** TravT has joined #openstack-relmgr-office05:37
*** TravT has quit IRC05:57
*** TravT has joined #openstack-relmgr-office06:21
*** jroll has quit IRC06:27
*** TravT has quit IRC06:30
*** TravT has joined #openstack-relmgr-office06:31
*** jroll has joined #openstack-relmgr-office06:33
*** armax has quit IRC06:38
ttxthingee: I'll probably let it live until Monday, I think we need to cut the library stable branches first06:54
ttxhopefully dhellmann will be back today and be able handle those06:55
*** HenryG has quit IRC07:21
ttx(quick note, I'm AFK today until ~ 1300 utc)07:31
*** HenryG has joined #openstack-relmgr-office08:12
*** HenryG has quit IRC08:16
*** dims_ has joined #openstack-relmgr-office08:30
*** HenryG has joined #openstack-relmgr-office08:30
*** dims_ has quit IRC08:34
*** dims_ has joined #openstack-relmgr-office09:32
*** dims_ has quit IRC09:36
*** dims_ has joined #openstack-relmgr-office09:50
dims_ttx: am picking up library release request https://review.openstack.org/#/c/224739/ and https://review.openstack.org/#/c/224812/ for keystone09:52
openstackgerritMerged openstack/releases: Release 1.1.0 of keystoneauth  https://review.openstack.org/22473910:47
openstackgerritMerged openstack/releases: Release 2.3.0 of keystonemiddleware  https://review.openstack.org/22481210:48
*** gordc has joined #openstack-relmgr-office11:41
*** bswartz has quit IRC12:30
*** mriedem has joined #openstack-relmgr-office12:41
dhellmannttx: back and catching up on scrollback12:44
dhellmanndims_: thanks for picking up the slack while I was out12:44
dims_yay, welcome back doug12:45
dhellmannlifeless: I may have been overly conservative with interpreting the semver rules, but if the minimum allowed version of a dep is raised then the new version of the lib isn't compatible with the old version any more, is it?12:46
dhellmanndims_, ttx, lifeless : where do we stand on library constraint updates?12:50
dhellmannhttps://review.openstack.org/#/q/project:openstack/requirements+is:open+branch:master+file:upper-constraints.txt,n,z12:51
dims_dhellmann: i haven't been tracking that12:54
dhellmannnp12:54
dhellmanndims_: did your fix for the - _ issue in glance-store land?13:01
dims_dhellmann: yes - https://review.openstack.org/#/c/224259/13:01
dhellmanndims_: does pip handle the safe_name() output correctly as a match? how about devstack, where I think we were filtering out LIBS_FROM_GIT from the constraints file?13:03
dims_dhellmann: pip does have a problem as lifeless mentions - "firstly, pip should have errored, because we had two constraints for the same thing"13:07
dhellmannI'm not sure we ever had 2 constraints, did we?13:07
dhellmannI thought it was just spelled wrong13:07
dims_dhellmann: what was happening was that edit-contraints was adding 'django_openstack_auth" with the file/path location even when there was already a django-openstack-auth line13:08
dhellmannah, I see. We also had an issue with glance_store being glance-store but I think that was done by hand13:09
dims_the review above, makes sure that both situations are handled13:09
*** dtantsur has joined #openstack-relmgr-office13:09
* dhellmann nods13:09
dtantsurdhellmann, o/13:09
dims_dhellmann: y13:09
dtantsurdhellmann, that's about https://review.openstack.org/#/c/223143/ I'm fine with not bumping g-r, but I see tripleoclient trying to use 1.2.0 features13:10
dhellmanndims_: the other side of the issue is that the code in devstack that tries to install libraries from git wasn't finding the name if it didn't match the output of "python setup.py --name" for the project in question13:10
dims_dhellmann: from a grep on my local directory it's just glance_store and django_openstack_auth with the unsafe names in their setup.cfg13:10
dhellmanndtantsur: are they already trying to use them, or do they have plans to use them in the near future?13:10
dhellmanndims_: that sounds right13:10
dtantsurdhellmann, there's a patch up, lemme find13:10
dhellmanndtantsur: if it is not already merged, I think it's too late13:11
dhellmanndims_, lifeless, ttx, sdague: it would be good to get some reviews on https://review.openstack.org/#/q/project:openstack/requirements+is:open+branch:master+file:upper-constraints.txt,n,z13:11
dtantsurdhellmann, ack, I'll communicate it to them. are you ok with upper-constraint change?13:12
dhellmanndims_, lifeless, ttx: since we're in freeze, I've been waiting for a second +2 on everything13:12
dhellmanndtantsur: yes, the upper-constraints.txt change is fine, but they need to be aware that does not mean they can land the patch and rely on the feature because their requirements will be wrong13:13
dtantsurdhellmann, yeah, I realize it. thanks!13:13
dhellmanndtantsur: ok, great, just making sure there's no mis-communication, thanks!13:14
dims_dhellmann: i can check and +1 in a few mins13:14
dtantsurdhellmann, updated the patch to only mention u-c13:14
dhellmanndtantsur: +2, ty13:15
dims_ouch don't know if any of those jobs use python-congressclient :( https://review.openstack.org/#/c/223130/13:16
dhellmannyeah, probably not13:16
dhellmannwe should at least have a job that tries to install everything listed, but I guess we don't right now13:17
dhellmannlifeless: given https://review.openstack.org/#/c/224943/ passed all its tests for a constrained version of lib that doesn't exist, I wonder if we're missing an "install everything" or at least "verify every constrained version is available" test. Didn't we used to have a test that installed all of the packages into a virtualenv?13:19
dhellmannor was that pbr?13:20
dims_don't believe i've run into that13:22
dims_jokke_: can you please respin https://review.openstack.org/#/c/223513/?13:22
* ttx is back13:22
dims_dhellmann: we should leave glance-store as-is ^^13:22
dims_hey ttx13:23
ttxdhellmann: reviewing13:23
dhellmanndims_: having the - is what broke devstack. It wasn't testing glance_store from source when we thought it was.13:23
dims_dhellmann: right, we fixed it now, so should be fine now13:24
dhellmanndims_: https://review.openstack.org/#/c/220648/13:24
dhellmanndims_: devstack is fixed, too? not just the constraints auto-update?13:24
dims_dhellmann: yes,13:25
dhellmanndims_: ack, good13:25
dims_the devstack job for django_openstack_auth against its master was the one that triggered this research13:25
ttxdhellmann: weird alpha ordering on https://review.openstack.org/#/c/223838 ... any chance you can update it ?13:26
dhellmanndims_: where was the change to fix devstack? does it use the edit-constraints code in the requirements repo?13:26
ttxsame for https://review.openstack.org/#/c/22383713:26
dhellmannttx: that's odd. I think that was generated by the edit-constraints, but I'll look13:27
dims_let me rethink this a bit...13:27
dhellmannttx: done13:28
jokke_dims_: looking13:28
dims_so here's my feeling, setup.cfg name should be safe (should not have _), then we will have the right names in g-r and upper-constraints. if setup.cfg name is bad. requirements and u-c should still have the safe names and devstack and constraints update should both work13:29
jokke_dims_: so will the tests now run if we keep it as glance-store?13:29
dhellmannjokke_: we're working through that right now13:30
dims_jokke_: we should try, want to spin another review to try that please?13:30
jokke_aha13:30
jokke_ok ... gimme a sec13:30
dims_thanks13:30
ttxdhellmann: same for https://review.openstack.org/#/c/22383713:30
jokke_ok, new one out there13:32
dhellmannttx: ahead of you13:32
dhellmanndims_: devstack has "glance_store" hard-coded as a name in inc/glance inside install_glance()13:33
*** dansmith is now known as superdan13:35
*** jhesketh has quit IRC13:44
*** jhesketh has joined #openstack-relmgr-office13:45
*** openstackgerrit has quit IRC13:46
*** openstackgerrit has joined #openstack-relmgr-office13:47
david-lyleafter many gate issues, django_openstack_auth is finally ready for release with 6e2cd8deffbd60074bf036c1cecdb280f281ccd413:48
david-lyleshould be 1.4.113:48
dhellmanndavid-lyle: great! please submit a release request to openstack/releases (see http://git.openstack.org/cgit/openstack/releases/tree/README.rst)13:48
david-lyledhellmann: yes, forgot13:49
dhellmanndavid-lyle: with the requirements updates I see, you want 1.5.0 to be consistent with how we've applied semver elsewhere13:50
dhellmanndavid-lyle: does "ab9678d 2015-08-25 11:17:35 -0600 Removing hack for python 2.6 support" remove python 2.6 support entirely?13:50
david-lyledhellmann: no that happened in 1.4.013:50
david-lylethis is just cleaning up some dead code13:51
david-lyles/this/that/13:51
dhellmanndavid-lyle: ok, that slipped past, it should have been a 2.0.0 release at that point :-(13:51
david-lylewe went from 1.3.* to 1.4.* on that one13:51
dhellmannright, but dropping a language level support is a backwards-incompatible change13:52
david-lyledhellmann: right13:53
dhellmannlooking at the list of changes in that release, I'm not sure I would have spotted it either http://logs.openstack.org/89/214689/5/check/gate-releases-tox-list-changes/16d1c09/console.html.gz13:54
dhellmanndavid-lyle: did 1.4.0 have a change that actually broke 2.6?13:54
david-lyledjango 1.8 no longer supports python 2.613:55
dhellmannok, so it was buried in one of the requirements updates?13:56
david-lyleessentially13:56
dhellmannok. not a terribly big deal, but that's the sort of thing to highlight in future releases13:56
david-lyledhellmann: I understand13:57
david-lyledhellmann: the 1.4.1 or 1.5? for this release?13:59
david-lyle*then13:59
dhellmannttx, lifeless, david-lyle: since we missed the 2.0 bump for 1.4, but it applies to this upcoming release as well, should we do 1.5 or 2.0?13:59
dhellmanndavid-lyle: it will be at least 1.513:59
dhellmanndims_: thoughts? ^^14:00
david-lylewell we went from 1.3. - 1.4 in Liberty for the django/python change14:00
david-lylebut we can move to 2.014:00
dhellmanndavid-lyle: yeah, the 1.3-1.4 change was a mistake, so I'm trying to decide if we correct it now for the stable branch release14:01
dhellmannI'm inclined to do it, but want to see what the others think14:01
david-lylesure14:01
dims_+1 to 2.014:02
openstackgerritBen Swartzlander proposed openstack/releases: Release of python-manilaclient 1.4.0  https://review.openstack.org/22515914:02
dhellmanndavid-lyle: why don't you go ahead and submit it as 2.0.0 and if the others think 1.5 is ok I'll edit your patch before cutting the release14:02
*** bswartz has joined #openstack-relmgr-office14:02
david-lyledhellmann: sounds good14:02
bswartzdhellmann: I just pushed https://review.openstack.org/225159 to do another manilaclient release for liberty -- it that going to be controversial or are you okay with it?14:03
dhellmannbswartz: as long as it's backwards-compatible, that should be ok. I think at this point we wouldn't want any other projects to raise their dependency on manilaclient unless the feature they need is already landed, but it doesn't seem like that would be possible.14:05
bswartzyeah no other projects will have a dependency on 1.4, but 1.4 should be the CLI version that ships with Liberty14:07
bswartzso the 1.4 release is where the manilaclient branch should be made14:08
openstackgerritDavid Lyle proposed openstack/releases: Release 2.0.0 of django_openstack_auth  https://review.openstack.org/22516714:08
dhellmannbswartz: ok, sounds good14:12
bswartzdhellmann: also I updated the requirements patch that bumps the upper-contraint on manilaclient to depend on the release of the mentioned version, so they can be workflowed in any order14:13
*** sigmavirus24_awa is now known as sigmavirus2414:14
dhellmannbswartz: sounds good.14:15
*** dims_ has quit IRC14:20
*** dims_ has joined #openstack-relmgr-office14:21
openstackgerritMerged openstack/releases: Release of python-manilaclient 1.4.0  https://review.openstack.org/22515914:21
*** Guest67 has joined #openstack-relmgr-office14:21
jokke_dims_: looks promising14:26
dims_jokke_: cool14:26
*** Guest67 is now known as nikhil_k14:28
*** jroll is now known as jroll|dupe14:31
*** jroll|dupe is now known as brickednick14:31
openstackgerritMerged openstack/releases: Release 2.0.0 of django_openstack_auth  https://review.openstack.org/22516714:31
*** david-lyle has quit IRC14:31
*** brickednick is now known as jroll14:31
* ttx looks up FFE/RC1 progress14:33
ttxnikhil_k, mestery, redrobot, morgan: ping me when around for a quick release status check14:34
ttxSlickNik: same ^14:35
*** nikhil has joined #openstack-relmgr-office14:35
nikhil_kttx: hi14:35
nikhil_khttps://launchpad.net/glance/+milestone/liberty-rc114:35
ttxnikhil: how are those last feature patches going on ?14:36
nikhil_kI got this approved14:36
nikhil_khttps://review.openstack.org/#/c/196240/14:36
nikhil_kthe other one is ready code wise14:36
ttxso one is in-flight, the other missing final review ?14:36
nikhil_kI will let you know (my) early Monday on the final call14:36
nikhil_kyes14:36
ttxok. You still need to refine the RC1 bug list so that it's truly release blockers. It feels like tere are a few nice-to-haves in there14:37
nikhil_kttx I got the artifacts people open a bug for each of their reviews14:38
ttxah, ok14:38
nikhil_kttx so basically, they all are waiting for one more +214:38
nikhil_kthe bug list is relatively small14:38
ttxsounds good14:38
ttxas long as we know what we are waiting for, it's good14:39
nikhil_kawesome14:39
nikhil_kttx we are targetting late Monday or Tuesday max, depending on how the gate behaves14:39
nikhil_kpeople are made aware of the same14:39
ttxwould be really good if all feature patches were in before end of week. We are already one week late for those14:40
ttxso final call on Monday morning sounds good.14:40
nikhil_kyeah, I hope for the same14:40
nikhil_kand I will send email asking for reviews so that people can them over/after weekend14:40
openstackgerritDoug Hellmann proposed openstack-infra/release-tools: add special case for django-openstack-auth  https://review.openstack.org/22518014:40
ttxthen it's fine if we spend a couple of days fixing the RC bugs14:40
ttxnikhil_k: thx14:40
nikhil_kthanks ttx14:40
ttxdhellmann: where do we stand wrt libraries and stable branches there ?14:43
ttxAll oslo libs are done I think14:43
dims_yep14:43
dhellmannttx: I haven't started cutting the others, yet. we're still getting release requests trickling in14:44
dhellmannand I want the constraints updates to go in before the branches14:44
ttxyeah. We still have FFEs running too, so we are not exactly on track14:44
ttxwhich makes me wonder about reducing that period by one week14:44
ttxmaking if we make a lot of noise about that one week reductyyion people will be more careful with deadlines (+ documentatin of the whole cycle)14:45
dhellmannyeah, my hope was that next cycle we would have to do less work chasing folks14:45
dhellmannagreed, having a more detailed cycle schedule with all the dates and implications listed, may help14:45
dhellmannI was also going to ask all PTLs for managed projects to ack all release-related announce emails on the -dev list so we know who's paying attention14:46
ttxdhellmann: it's the first cycle that I leave FFEs entirely up to PTLs, and some of them were stricter than I was, but about half of them were a lot looser and we still have standing FFes 3 weeks after FF14:47
dhellmann:-/14:47
*** TravT_ has joined #openstack-relmgr-office14:47
*** TravT has quit IRC14:47
dims_ouch14:48
*** TravT_ has quit IRC14:48
ttxif we still have some next week, we might need to use a bit of our influence to ask them to yank them14:48
*** TravT has joined #openstack-relmgr-office14:48
dhellmannif we're already behind, should we wait until next week?14:49
dhellmannI've rejected one or two requirements updates to enable features already14:49
ttxdhellmann: I think we are still OK with all RC1s before end of month14:49
dhellmannok14:49
ttxdue to small amounts of RC bugs14:49
ttxbut I epect a lot of respins as people take those late features out for real -life testing14:50
ttxbasically we are not (yet) jeopardizing final release date. We are reducing the quality of the end release.14:50
ttx(by accepting features late and reducing testing time)14:50
*** TravT has quit IRC14:53
ttxdhellmann: I'm approving tha last two release tag changes (keystone and deb-packaging)14:53
*** TravT has joined #openstack-relmgr-office14:53
ttxwe can always fix them later14:53
*** nikhil has quit IRC14:54
*** TravT_ has joined #openstack-relmgr-office14:59
*** TravT_ has quit IRC14:59
*** TravT has quit IRC14:59
*** TravT has joined #openstack-relmgr-office15:00
*** armax has joined #openstack-relmgr-office15:00
*** dims_ is now known as dimsum__15:03
dhellmannttx: ++15:04
*** TravT has quit IRC15:05
*** TravT has joined #openstack-relmgr-office15:05
dhellmannttx, dimsum__, lifeless : I've updated https://review.openstack.org/#/c/223596/, which updates several constraints at once15:08
ttx+215:09
dimsum__+115:37
*** david-lyle has joined #openstack-relmgr-office15:43
mesteryttx: o/15:45
mesteryttx: Still waiting on 4 BPs, but I've got updates on them for you15:45
ttxsure, fire15:47
mesteryhttps://blueprints.launchpad.net/neutron/+spec/lbaas-ref-octavia: Waiting on a single patch to merge in the neutron-lbaas repo which enables octavia by default in devstack15:47
mesteryhttps://blueprints.launchpad.net/neutron/+spec/wsgi-pecan-switch: Two patches to merge, then fold back to master, then it's done15:47
mesteryhttps://blueprints.launchpad.net/neutron/+spec/external-dns-resolution: One patch left, if it's not in by Monday it's out of Liberty15:47
mesteryhttps://blueprints.launchpad.net/neutron/+spec/neutron-flavor-framework-templates: This one is the most at risk, and likely won't make it. I'm following up with the owners today15:48
*** mriedem is now known as mriedem_lunch15:48
ttxyeah, if it's not in on Monday I think it's safer to defer at this stage15:51
mesteryAck15:55
mesteryWe'll work to get these landed today and over the weekend go from there.15:55
mesteryMonday things will be a lot more clear15:55
mesteryThanks ttx!15:55
ttxcool, thx15:55
*** dtantsur is now known as dtantsur|afk16:01
dhellmannI keep being surprised by the number of libraries we have now. This script to check the current versions of libs against constraints takes a long time to run.16:18
*** david-lyle has quit IRC16:19
*** david-lyle has joined #openstack-relmgr-office16:26
*** david-lyle has quit IRC16:31
mesterydhellmann: #winning? :)16:39
openstackgerritDoug Hellmann proposed openstack-infra/release-tools: Show output of zuul-cloner  https://review.openstack.org/22522916:40
openstackgerritDoug Hellmann proposed openstack-infra/release-tools: add a tool to compare the constraints file with existing releases  https://review.openstack.org/22523016:40
dhellmannmestery: something like that :-)16:41
mesterylol16:41
mestery:)16:41
dimsum__we so sooooo modular now :)16:50
dimsum__we should turn each library into a micro service16:50
*** david-lyle has joined #openstack-relmgr-office16:57
*** david-lyle has quit IRC17:44
morganttx: here17:54
morganttx: sorry for the delay. Missed the ping earlier17:54
*** bswartz has quit IRC17:58
*** TravT has quit IRC18:09
*** TravT_ has joined #openstack-relmgr-office18:09
*** david-lyle has joined #openstack-relmgr-office18:17
*** TravT_ has quit IRC18:35
*** TravT has joined #openstack-relmgr-office18:36
*** TravT has quit IRC18:38
*** TravT has joined #openstack-relmgr-office18:38
*** david-lyle has quit IRC19:11
*** david-lyle has joined #openstack-relmgr-office19:11
*** david-lyle has quit IRC19:19
openstackgerritDoug Hellmann proposed openstack-infra/release-tools: add a tool to compare the constraints file with existing releases  https://review.openstack.org/22523019:27
lifelessttx: http://logs.openstack.org/57/221157/10/check/gate-tempest-dsvm-full/18eb440/logs/devstacklog.txt.gz#_2015-09-18_13_51_46_902 is worrying19:30
lifelessits my weekend but I'm going to send a mail19:30
*** TravT_ has joined #openstack-relmgr-office19:42
*** TravT has quit IRC19:44
*** TravT_ is now known as TravT19:50
*** mriedem_lunch is now known as mriedem20:06
*** david-lyle has joined #openstack-relmgr-office20:10
*** mriedem has left #openstack-relmgr-office20:12
*** bnemec has quit IRC20:15
*** bnemec has joined #openstack-relmgr-office20:15
*** gordc has quit IRC20:24
dimsum__lifeless: dhellmann: what is the version for say Nova when we do make the liberty release?20:48
dhellmanndimsum__: each project will have its own version, depending on how many releases have been done previously20:48
dhellmanndimsum__: nova will be 12.x.y20:49
dimsum__dhellmann: ack thanks20:59
dhellmannttx: we have patches in flight to update the constraints for all libraries that were out of date. Most are in https://review.openstack.org/223596 but there are one or two individual patches still.21:08
dhellmannttx: mordred is working to resolve the osc/occ issue with a patch to occ so we can update its constraint as well, and that will happen when I do the next release21:08
dhellmannttx: I have added to my series of release-tools updates in https://review.openstack.org/#/q/project:openstack-infra/release-tools+is:open+topic:fix-alpha-tag-handling,n,z21:10
dhellmannttx: and with that status update, I'm going to start my weekend21:10
*** dims_ has joined #openstack-relmgr-office21:29
*** dimsum__ has quit IRC21:32
dims_good night dhellmann get some rest :)21:35
openstackgerritMatt Riedemann proposed openstack/releases: Juno: Release python-ceilometerclient 1.0.15  https://review.openstack.org/22533921:38
*** david-lyle has quit IRC22:06
*** sigmavirus24 is now known as sigmavirus24_awa22:32
*** openstackgerrit has quit IRC22:46
*** openstackgerrit has joined #openstack-relmgr-office22:46
*** dims_ has quit IRC22:56
*** dimsum__ has joined #openstack-relmgr-office23:23
*** nikhil_k has quit IRC23:29

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!