mozilla :: #releng

19 Apr 2017
06:38whimbooHi. Did we already stop updates for Aurora?
06:41whimboohm, looks like it was missed to update mozmill-ci for teh auroratest channel
06:48FlorinMezeiwhimboo: I was supposed to do that, but thought that it didn't matter anymore since Aurora is going away
06:48whimbooFlorinMezei: as long as we serve updates we have to test them
06:48whimboowhen funsize stops to produce the partials we cannot run tests anymore
06:48whimboothat's our stop gap
06:49FlorinMezeigiven that the merge was yesterday I don't think we'll have any more builds
06:49FlorinMezeior am I wrong?
06:50whimbooi don't know
06:50FlorinMezei aurora nightly updates are locked to current latest buildid: o Fennec-mozilla-aurora-nightly-20170418074655 o Firefox-mozilla-aurora-nightly-20170418074700 o Both build ids point to https://treeherder.mozilla.org/#/jobs?repo=mozilla-aurora&revision=105e456d811b8c4616de688e9afdcf0af620b80b
06:51FlorinMezeifrom the merge email yesterday
06:51whimboosfraser: ^ does it mean we stop funsize from creating the partials now?
06:52whimbooor if no nightly is build we won't run those jobs anyway
06:52whimboobut there might be a chemspill like release if it would become necessary while transition users over to beta
06:53whimbooFlorinMezei: so I would suggest we keep our tests active for now, until the transition has been done
06:53FlorinMezeiwhimboo: got it... should I then set auroratest?
06:53FlorinMezeior have you done that?
06:57whimbooFlorinMezei: i already did
06:58FlorinMezeiok, I guess that if the aurora channel is open again I'll have to change it back to "aurora"
06:58FlorinMezeiso I'll look out for that
07:05sfraserwhimboo: if no nightly, no partials will be generated
08:07travis-cibuild-tools#1744 (master - a4c5ecb : Johan Lorenzo): The build passed. (https://travis-ci.org/mozilla/build-tools/builds/223459280)
08:30travis-cibuild-puppet#1215 (master - 597ed44 : Dragos Crisan): The build passed. (https://travis-ci.org/mozilla/build-puppet/builds/223465581)
08:58travis-cibuild-puppet#1216 (production - a48dda4 : Andrei Obreja): The build passed. (https://travis-ci.org/mozilla/build-puppet/builds/223472899)
13:02travis-cibuild-puppet#1217 (master - 953b60b : Dave House): The build passed. (https://travis-ci.org/mozilla/build-puppet/builds/223545881)
13:17travis-cibuild-puppet#1218 (production - 039e3ad : Dave House): The build passed. (https://travis-ci.org/mozilla/build-puppet/builds/223551547)
19:53travis-cibuild-buildbot-configs#2549 (production - 5ce9b0a : ffxbld): The build passed. (https://travis-ci.org/mozilla-releng/build-buildbot-configs/builds/223698387)
19:53travis-cibuild-buildbotcustom#1031 (production-0.8 - 8fe603c : ffxbld): The build passed. (https://travis-ci.org/mozilla-releng/build-buildbotcustom/builds/223698396)
19:54travis-cibuild-buildbot-configs#2548 (master - 989e2aa : ffxbld): The build passed. (https://travis-ci.org/mozilla-releng/build-buildbot-configs/builds/223698381)
19:55travis-cibuild-tools#1745 (master - 492af47 : ffxbld): The build passed. (https://travis-ci.org/mozilla/build-tools/builds/223698418)
21:10mbrandtA couple our go-bouncer tests flagged a potential error. Several of the winxp tests are failing, I would expect 52.1.0esr to be served up (https://product-details.mozilla.org/1.0/firefox_versions.json) but 52.0.2esr is being served for these aliases https://gist.github.com/ed66b50d2707dccd14ac8e2af89d0d00
21:11mbrandtrail: ^ not sure if you're the person to ping
21:11raillooking
21:12railprobably https://bugzilla.mozilla.org/show_bug.cgi?id=1340263 is not yet deployed, I'll update the aliases manually
21:13mbrandtthanks rail. appreciated
21:15railnp
21:15railmbrandt: it should be fine now
21:15rstrongrail: do you know if updates on 51 have been disabled while we are trying to release 53?
21:15rstrongat any point that is
21:16rstrongperhaps jlund^^
21:16railhmm....
21:16mbrandtrail: manually, it looks spot on. Kicking off our automated checks now
21:17railI don't recall anything 51 specific...
21:18rstrongrail: not specific to 51.
21:18rstronghave updates been disabled at all?
21:19railah
21:19rstrongThis has happened on previous releases and I'm getting pinged and bug reports
21:19raillet me check
21:25railrstrong: I think I see what you mean. ATM we serve Firefox-53.0-build6 at 25% and Firefox-51.0.1-build3 at 75% (fallback). probably we either didn't update the fallback or something went wrong
21:25railbhearsum: do you know if we set the fallback manually or automatic?
21:26bhearsumrail: it is not set automatically yet
21:27bhearsumit's something that releaseduty should be doing (not trying to throw them under the bus, though)
21:27railI' going to set set it to something fresher
21:27akiit's not doc'ed
21:28akii just added the 25% line; could someone make sure whatever we need to be doing is doc'ed here? https://github.com/mozilla/releasewarrior/blob/master/how-tos/relpro.md#how-2
21:28bhearsumi'll update that
21:28railI'll update the rule to fallback to Firefox-52.0.2-build1
21:29jlundUntil automated, I suppose this should be a step in rel warrior?
21:29bhearsumyes
21:29bhearsumpart of the "publish" step
21:29akisure, or at least in the docs for publish
21:30bhearsumit will get automated as part of https://bugzilla.mozilla.org/show_bug.cgi?id=1347212
21:30bhearsumwhich probably won't be until after dawn is completed
21:30raildone
21:30railinteresting, it didn't conflict with the scheduled change
21:31bhearsumthat's a 3 way merge
21:31railooh
21:31bhearsumit only conflicts if both are changing the same field
21:31railI see
21:31railsweet
21:32akidoes anyone have access to the ship-it db?
21:32akinm :)
21:36bhearsumhttps://github.com/mozilla/releasewarrior/pull/73
21:38akithat's essentially the latest shipped release?
21:39bhearsumyep
21:39aki
21:39bhearsumas long as you do it prior to unblocking the human decision task
21:40bhearsumthanks
21:40bhearsumsorry about that - this being missed is totally not your fault
21:41akinw, now we have it documented
20 Apr 2017
No messages
   
Last message: 8 days and 18 hours ago