+++ This bug was initially created as a clone of Bug #1253749 +++ Similar to Bug #1253749, but for esr45

Review commit: https://reviewboard.mozilla.org/r/52475/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/52475/

Comment on attachment 8752226 [details] MozReview Request: Bug 1253757 - Branch specifics for relpro on esr45 r=jlund a=release DONTBUILD https://reviewboard.mozilla.org/r/52475/#review49582 looks good. we plan on trying this in staging? if so, we will need staging/dev configs for updates/postrelease etc ::: testing/mozharness/configs/builds/branch_specifics.py:196 (Diff revision 1) > + 'branch_uses_per_checkin_strategy': True, > + 'use_branch_in_symbols_extra_buildid': False, > + 'stage_server': 'upload.ffxbld.productdelivery.prod.mozaws.net', > + 'platform_overrides': { > + 'linux': { > + 'src_mozconfig': 'browser/config/mozconfigs/linux32/release', granted esr45 ci builds will be so rare that they will probably be clobber jobs, should we force or even explicitly say we want these to be clobbers via: 'force_clobber': True, ?

Comment on attachment 8752226 [details] MozReview Request: Bug 1253757 - Branch specifics for relpro on esr45 r=jlund a=release DONTBUILD Review request updated; see interdiff: https://reviewboard.mozilla.org/r/52475/diff/1-2/

(In reply to Jordan Lund (:jlund) from comment #2) > granted esr45 ci builds will be so rare that they will probably be clobber > jobs, should we force or even explicitly say we want these to be clobbers > via: 'force_clobber': True, ? Good catch. For some reason I missed that...

Comment on attachment 8752226 [details] MozReview Request: Bug 1253757 - Branch specifics for relpro on esr45 r=jlund a=release DONTBUILD https://reviewboard.mozilla.org/r/52475/#review49638

Comment on attachment 8752226 [details] MozReview Request: Bug 1253757 - Branch specifics for relpro on esr45 r=jlund a=release DONTBUILD https://hg.mozilla.org/integration/mozilla-inbound/rev/afda2f94c0c1 aurora: todo https://hg.mozilla.org/releases/mozilla-beta/rev/1f0711112a6f https://hg.mozilla.org/releases/mozilla-release/rev/9a8bf4390f5d https://hg.mozilla.org/releases/mozilla-esr45/rev/3e1f6768ee3a

Aurora was closed, need to push it there too.

Morphing the title to keep everything under one bug

Review commit: https://reviewboard.mozilla.org/r/53246/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/53246/

Comment on attachment 8753384 [details] MozReview Request: Bug 1253757 - Add esr update verify configs, delete unused mozRelease configs. r=Callek https://reviewboard.mozilla.org/r/53246/#review50030

Comment on attachment 8753384 [details] MozReview Request: Bug 1253757 - Add esr update verify configs, delete unused mozRelease configs. r=Callek https://hg.mozilla.org/build/tools/rev/b471ad9770be

Comment on attachment 8752226 [details] MozReview Request: Bug 1253757 - Branch specifics for relpro on esr45 r=jlund a=release DONTBUILD https://hg.mozilla.org/releases/mozilla-aurora/rev/049fb75980cf

* Disable ESR nightlies * Enable mozhareness for CI builds * Add build promotion specific configs Review commit: https://reviewboard.mozilla.org/r/53554/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/53554/

Comment on attachment 8753948 [details] [diff] [review] release-runner-esr45.diff Review of attachment 8753948 [details] [diff] [review]: ----------------------------------------------------------------- ::: lib/python/kickoff/__init__.py @@ +12,2 @@ > > # temporary regex to filter out anything but mozilla-beta and mozilla-release and esr45!

Comment on attachment 8752226 [details] MozReview Request: Bug 1253757 - Branch specifics for relpro on esr45 r=jlund a=release DONTBUILD https://reviewboard.mozilla.org/r/52475/#review50460 ::: testing/mozharness/configs/releases/updates_firefox_esr45.py:30 (Diff revision 2) > + "patcher_config": "mozEsr45-branch-patcher2.cfg", > + "update_verify_channel": "esr-localtest", > + "mar_channel_ids": [], > + "channel_names": ["esr", "esr-localtest", "esr-cdntest"], > + "rules_to_update": ["esr45-cdntest", "esr45-localtest"], > + "publish_rules": ["esr"], just noticing, should this be the alias "esr45"?

Comment on attachment 8753904 [details] MozReview Request: Bug 1253757 - Switch mozilla-esr45 to release build promotion r=jlund https://reviewboard.mozilla.org/r/53554/#review50456 we are going to need to tweak release-runner.py if you haven't done so already in another bug. see postrelease_* comments in line. also, we will likely need to address this hardcoded block logic: * http://hg.mozilla.org/build/tools/file/default/buildfarm/release/release-runner.py#l373 * http://hg.mozilla.org/build/tools/file/default/buildfarm/release/release-runner.py#l371 :( I wish every release variant had all their items consolidated in one place like: https://github.com/mozilla/releasetasks/tree/master/releasetasks/release_configs ::: mozilla/config.py:2766 (Diff revision 1) > +### Release Promotion > +BRANCHES['mozilla-esr45']['enable_release_promotion'] = True > +# used by process/release.py > +BRANCHES['mozilla-esr45']['tuxedoServerUrl'] = "https://bounceradmin.mozilla.com/api" > +BRANCHES['mozilla-esr45']['bouncer_submitter_config'] = { > + "firefox": "releases/bouncer_firefox_esr.py", these are coming next? ::: mozilla/config.py:2792 (Diff revision 1) > +BRANCHES['mozilla-esr45']['postrelease_version_bump_enabled'] = True > +BRANCHES['mozilla-esr45']['postrelease_bouncer_aliases_enabled'] = True I think given that we want a graph2 for esr releases, we will want these set to False. Alternatively if we only want to do 2 graphs for certain esr releases, we would have to teach release-runner to override these bools like we do for RC releases. e.g. say if if we want to have a graph2 for minor esr dot releases but not for major esr releases

https://reviewboard.mozilla.org/r/53554/#review50456 > these are coming next? forgot I reviewed these :)

Comment on attachment 8753904 [details] MozReview Request: Bug 1253757 - Switch mozilla-esr45 to release build promotion r=jlund Let's try again. :) This should just work (TM), assuming we are ok with having graph2-only tasks in graph1 as well. TBH, I don't want to touch the is_release_candidate() logic and have something similar for ESR. This approached worked fine in bug 1276250.

Comment on attachment 8753904 [details] MozReview Request: Bug 1253757 - Switch mozilla-esr45 to release build promotion r=jlund https://reviewboard.mozilla.org/r/53554/#review52756

Comment on attachment 8753904 [details] MozReview Request: Bug 1253757 - Switch mozilla-esr45 to release build promotion r=jlund https://hg.mozilla.org/build/buildbot-configs/rev/d3363d0677d3

Comment on attachment 8753946 [details] [diff] [review] release-runner-esr38.diff https://hg.mozilla.org/build/tools/rev/5c5787382b88

Comment on attachment 8753948 [details] [diff] [review] release-runner-esr45.diff https://hg.mozilla.org/build/tools/rev/e7fdef596219

In production: https://hg.mozilla.org/build/buildbot-configs/rev/d3363d0677d3

I also pushed in-tree version fix: https://hg.mozilla.org/releases/mozilla-esr45/rev/6f20e58674c6

time to clean up! (default branch)

old-release-runner branch

Comment on attachment 8759308 [details] [diff] [review] default_patterns.diff Review of attachment 8759308 [details] [diff] [review]: ----------------------------------------------------------------- wooohooo

Comment on attachment 8759308 [details] [diff] [review] default_patterns.diff remote: https://hg.mozilla.org/build/tools/rev/2caef2b63d0e remote: https://hg.mozilla.org/build/tools/rev/322553bf6fec

Done!

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