1319436 - Tracking bug for 2017-06-13 migration work

Status: RESOLVED FIXED

(Release Engineering :: Release Requests, defect, P1)

Description

[Rail Aliiev [:rail]]

+++ This bug was initially created as a clone of Bug #1318919 +++

Comment 1

[Mihai Tabara [:mtabara]⌚️GMT]

Attachment: Bug 1319436 - bump release to 54.

Review commit: https://reviewboard.mozilla.org/r/145014/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/145014/

Comment 2

[Aki Sasaki (not active)]

Comment on attachment 8873633 [details]
Bug 1319436 - bump release to 54.

https://reviewboard.mozilla.org/r/145014/#review148972

Comment 3

[Mihai Tabara [:mtabara]⌚️GMT]

Attachment: Bug 1319436 - bump the remaining branches. r=jlorenzo

Comment 4

[Johan Lorenzo [:jlorenzo]]

Comment on attachment 8873822 [details] [diff] [review]
Bug 1319436 - bump the remaining branches. r=jlorenzo

I don't think we want to bump mozilla-aurora, because [1] will probably remain stuck on 54 forever. I guess the same applies to comm-aurora.

Apart from that, lgtm!

[1] https://dxr.mozilla.org/mozilla-aurora/source/browser/config/version.txt

Comment 5

[Mihai Tabara [:mtabara]⌚️GMT]

(In reply to Johan Lorenzo [:jlorenzo] from comment #4)
> Comment on attachment 8873822 [details] [diff] [review]
> Bug 1319436 - bump the remaining branches. r=jlorenzo
> 
> I don't think we want to bump mozilla-aurora, because [1] will probably
> remain stuck on 54 forever. I guess the same applies to comm-aurora.
> 
> Apart from that, lgtm!
> 
> [1] https://dxr.mozilla.org/mozilla-aurora/source/browser/config/version.txt

Thanks Johan!
Let's ask and confirm with :aki as well:
@aki: quick question for next week merge: should we freeze aurora to 54, drop it completely (might need to alter/delete other configs in this case I suppose) or bump it along with beta version?

Comment 6

[Mihai Tabara [:mtabara]⌚️GMT]

Comment on attachment 8873633 [details]
Bug 1319436 - bump release to 54.

Bug 1319436 - bump release to 54. r=aki
https://hg.mozilla.org/build/buildbot-configs/rev/748e07c05bb6
https://hg.mozilla.org/build/buildbot-configs/rev/cd5d62acf8b6

Comment 7

[Mihai Tabara [:mtabara]⌚️GMT]

(In reply to Mihai Tabara [:mtabara]⌚️GMT from comment #6)
> Comment on attachment 8873633 [details]
> Bug 1319436 - bump release to 54.
> 
> Bug 1319436 - bump release to 54. r=aki
> https://hg.mozilla.org/build/buildbot-configs/rev/748e07c05bb6
> https://hg.mozilla.org/build/buildbot-configs/rev/cd5d62acf8b6

I forgot we have automatic tests on bb-configs and merged before results arrived.
And sadly the push to default fails for the following error: (more at https://travis-ci.org/mozilla-releng/build-buildbot-configs/jobs/239519290 ). I suspect it's not related to my change itself but something else in one of tools/buildbotcustom/buildbot/braindump changed in the past five days. Investigating now ...


Finished processing dependencies for buildbot===0.8.2-hg-7f46f1a5f47e-production-0.8

outputting to allthethings.json

Checking 18 masters...

bm01-tests1-linux32 bm51-tests1-linux64 bm69-tests1-windows bm70-build1 bm75-try1 bm81-build_scheduler bm81-tests_scheduler bm103-tests1-linux bm106-tests1-macosx bm109-tests1-windows bm01-tests1-linux32-universal bm51-tests1-linux64-universal bm69-tests1-windows-universal bm70-build1-universal bm75-try1-universal bm103-tests1-linux-universal bm106-tests1-macosx-universal bm109-tests1-windows-universal

skipping universal master bm01-tests1-linux32-universal

skipping universal master bm70-build1-universal

skipping universal master bm106-tests1-macosx-universal

skipping universal master bm51-tests1-linux64-universal

skipping universal master bm69-tests1-windows-universal

skipping universal master bm75-try1-universal

skipping universal master bm109-tests1-windows-universal

skipping universal master bm103-tests1-linux-universal

2017-06-05 09:24:06,398 - loading test-output/bm109-tests1-windows/master.cfg

2017-06-05 09:24:06,411 - loading test-output/bm51-tests1-linux64/master.cfg

2017-06-05 09:24:06,412 - loading test-output/bm01-tests1-linux32/master.cfg

2017-06-05 09:24:06,415 - loading test-output/bm69-tests1-windows/master.cfg

2017-06-05 09:24:06,416 - loading test-output/bm70-build1/master.cfg

2017-06-05 09:24:06,418 - loading test-output/bm81-build_scheduler/master.cfg

2017-06-05 09:24:06,419 - loading test-output/bm103-tests1-linux/master.cfg

2017-06-05 09:24:06,420 - loading test-output/bm106-tests1-macosx/master.cfg

2017-06-05 09:24:06,432 - loading test-output/bm81-tests_scheduler/master.cfg

2017-06-05 09:24:06,432 - loading test-output/bm75-try1/master.cfg

2017-06-05 09:24:07,504 - Couldn't load test-output/bm81-tests_scheduler/master.cfg

Traceback (most recent call last):

  File "/home/travis/build/mozilla-releng/build-buildbot-configs/.tox/braindump/buildbot-related/dump_master_json.py", line 113, in dump_master

    c = loadMaster(path)

  File "/home/travis/build/mozilla-releng/build-buildbot-configs/.tox/braindump/buildbot-related/dump_master_json.py", line 26, in loadMaster

    execfile(path, g, g)

  File "/home/travis/build/mozilla-releng/build-buildbot-configs/test-output/bm81-tests_scheduler/master.cfg", line 7, in <module>

    import config

  File "/home/travis/build/mozilla-releng/build-buildbot-configs/test-output/bm81-tests_scheduler/config.py", line 2636, in <module>

    loadSkipConfig(BRANCHES, "desktop")

  File "/home/travis/build/mozilla-releng/build-buildbot-configs/test-output/bm81-tests_scheduler/config_seta.py", line 165, in loadSkipConfig

    define_configs(b, platforms, BRANCHES)

  File "/home/travis/build/mozilla-releng/build-buildbot-configs/test-output/bm81-tests_scheduler/config_seta.py", line 145, in define_configs

    platform = seta_platforms[p][0]

KeyError: 'Windows 10 64-bit'

Comment 8

[Mihai Tabara [:mtabara]⌚️GMT]

I could revert my patch and re-merge to production but that wouldn't help. I already tested against https://github.com/mozilla-releng/build-buildbot-configs/pull/11 and it doesn't solve the problem. It's infra related. Something changed in the past few days.

On another note, I *think* the change I landed is safe enough ATM in production that doesn't need backing it up.

Comment 9

[Mihai Tabara [:mtabara]⌚️GMT]

(In reply to Mihai Tabara [:mtabara]⌚️GMT from comment #8)
> I could revert my patch and re-merge to production but that wouldn't help. I
> already tested against
> https://github.com/mozilla-releng/build-buildbot-configs/pull/11 and it
> doesn't solve the problem. It's infra related. Something changed in the past
> few days.
> 
> On another note, I *think* the change I landed is safe enough ATM in
> production that doesn't need backing it up.

Status update: bug 1370182 tracked the issue with buildbot-configs tests bustage. 
All good and green now, resuming to mergeduty steps.

Comment 10

[Aki Sasaki (not active)]

Aiui, we should remove all automation references to aurora.

Comment 11

[Mihai Tabara [:mtabara]⌚️GMT]

(In reply to Aki Sasaki [:aki] from comment #10)
> Aiui, we should remove all automation references to aurora.

Okay, sgtm. I've added an item to discuss more of this during post-mortem.

Comment 12

[Mihai Tabara [:mtabara]⌚️GMT]

Comment on attachment 8873822 [details] [diff] [review]
Bug 1319436 - bump the remaining branches. r=jlorenzo

* munged a bit the patch to fix aurora to 54 to comemorate Dawn

Bug 1319436 - stabilizing branches merge day. r=jlorenzo
https://hg.mozilla.org/build/buildbot-configs/rev/45a1996a9a6d
https://hg.mozilla.org/build/buildbot-configs/rev/332d994a305d

Comment 13

[Mihai Tabara [:mtabara]⌚️GMT]

During l10n bumper step today, the script had two minor fails. Not worth filing a separate bug for that so I'll just comment here:

1. IOError: Can't find configs/l10n_bumper/mozilla-beta.py in ['.', '/builds/l10n-bumper/mozharness/scripts/../configs', '/builds/l10n-bumper/mozharness/scripts/../../configs']! by following the `l10n bumper` part from https://github.com/mozilla/releasewarrior/blob/master/how-tos/merge-day.md#merge-central-to-beta.

I'll address that changing the command to `mozharness/scripts/l10n_bumper.py -c configs/l10n_bumper/mozilla-beta.py --ignore-closed-tree` in a releasewarrior PR with other misc small fixes.

2.   File "mozharness/configs/l10n_bumper/mozilla-beta.py", line 11, in <module>
    "gecko_pull_url": "https://hg.mozilla.org/{}".format(MULTI_REPO),
ValueError: zero length field name in format

This is because https://hg.mozilla.org/releases/mozilla-beta/file/tip/testing/mozharness/configs/l10n_bumper/mozilla-beta.py#l11 lacks the positional arguments. In Python <= 2.6+ these were mandatory while after >= 2.7 these became optional. Sadly the BB masters run 2.6.6 python so this script fails. There are three ways to solve this:

a) we directly tweak locally the `configs/l10n_bumper/mozilla-beta.py#l11` mozharness file under under buildbot-master01.bb.releng.use1.mozilla.com for future use. The mozharness there is a one-time download so it's not version controlled. We update once and never deal with this again. 
b) we patch up https://hg.mozilla.org/releases/mozilla-beta/file/tip/testing/mozharness/configs/l10n_bumper/mozilla-beta.py#l11 to be `{0}` (compatible with both 2.6 and 2.7) in both ocurrences and land/uplift + redownload the mozharness part again on buildbot-master01.bb.releng.use1.mozilla.com 
c) we update python on BM :)

:aki - which one do you prefer?

Comment 14

[Aki Sasaki (not active)]

How about using /tools/python/bin/python which is 2.7?

Comment 15

[Mihai Tabara [:mtabara]⌚️GMT]

(In reply to Aki Sasaki [:aki] from comment #14)
> How about using /tools/python/bin/python which is 2.7?
Even better, thanks! Addressed here: https://github.com/mozilla/releasewarrior/pull/88

Comment 16

[Mihai Tabara [:mtabara]⌚️GMT]

We're far deep into 55 beta cycle and have no deps left from migration so I think it's safe to close this.

Comment 17

[Firefox Bug Husbandry Bot]

Removing leave-open keyword from resolved bugs, per :sylvestre.