Closed Bug 1039559 Opened 10 years ago Closed 10 years ago

No updates served for mozilla-esr31 nightly builds

Categories

(Release Engineering :: General, defect)

defect
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: whimboo, Assigned: bhearsum)

References

Details

(Whiteboard: [mozmill])

Attachments

(3 files)

As seen today via our Mozmill update tests, there are no updates served for nightly builds on the mozilla-esr31 update channel. Trying to do so manual also fails. Can we please check what's going on here?
There are no update snippets available as also seen in the log below: https://aus3.mozilla.org/update/3/Firefox/31.0/20140715000501/Linux_x86_64-gcc3/en-US/nightly-esr31/Linux%203.11.0-15-generic%20%28GTK%202.24.20%29/default/default/update.xml?force=1 Log output: 02:57:57 *** AUS:SVC gCanCheckForUpdates - able to check for updates 02:57:57 *** AUS:SVC gCanApplyUpdates - testing write access /home/mozauto/jenkins/workspace/mozilla-esr31_update/data/binary/firefox/update.test 02:57:57 *** AUS:SVC gCanApplyUpdates - able to apply updates 02:57:57 *** AUS:SVC UpdateService:removeDownloadListener - no downloader! 02:57:57 *** AUS:SVC Checker: checkForUpdates, force: true 02:57:57 *** AUS:SVC getLocale - getting locale from file: resource://gre/update.locale, locale: en-US 02:57:57 *** AUS:SVC Checker:getUpdateURL - update URL: https://aus3.mozilla.org/update/3/Firefox/31.0/20140715000501/Linux_x86_64-gcc3/en-US/nightly-esr31/Linux%203.11.0-15-generic%20(GTK%202.24.20)/default/default/update.xml?force=1 02:57:57 *** AUS:SVC Checker:checkForUpdates - sending request to: https://aus3.mozilla.org/update/3/Firefox/31.0/20140715000501/Linux_x86_64-gcc3/en-US/nightly-esr31/Linux%203.11.0-15-generic%20(GTK%202.24.20)/default/default/update.xml?force=1 02:57:57 *** AUS:SVC Checker:onLoad - request completed downloading document 02:57:57 *** AUS:SVC Checker:onLoad - number of updates available: 0 02:58:02 ERROR | Test Failure | { 02:58:02 "exception": { 02:58:02 "message": "An update has been found.", 02:58:02 "lineNumber": 27, 02:58:02 "name": "TimeoutError", 02:58:02 "fileName": "resource://mozmill/modules/errors.js" 02:58:02 } 02:58:02 }
http://mxr.mozilla.org/mozilla/source/webtools/aus/xml/inc/config-dist.php#440 needs to be updated with the new branches and deployed.
Blocks: 1012647
Comment on attachment 8457001 [details] [diff] [review] Bug 1039559 - No updates served for mozilla-esr31 nightly builds.patch I'm going to land this for Massimo
Attachment #8457001 - Flags: review?(rail) → review+
Comment on attachment 8457001 [details] [diff] [review] Bug 1039559 - No updates served for mozilla-esr31 nightly builds.patch rail@gueuze [13:07:36] ~/work/mozilla/build/aus/xml/inc ❯ cvs commit -m "Bug 1039559 - No updates served for mozilla-esr31 nightly builds. p=mgerva,r=rail" Checking in config-dist.php; /cvsroot/mozilla/webtools/aus/xml/inc/config-dist.php,v <-- config-dist.php new revision: 1.303; previous revision: 1.302 done rail@gueuze [13:08:34] ~/work/mozilla/build/aus/xml/inc ❯ cvs tag -F AUS2_PRODUCTION T config-dist.php
Attachment #8457001 - Flags: checked-in+
Depends on: 1039614
Oooh, fun edgecase. These are nightly builds, so should be on aus4.m.o (aka Balrog, I see someone added the rule for nightly-esr24 so it works). But the app is querying using aus3.m.o, which is what we want for the release builds. Bug 1027608 will change this, if we land it there.
(In reply to Nick Thomas [:nthomas] from comment #6) > Oooh, fun edgecase. These are nightly builds, so should be on aus4.m.o (aka > Balrog, I see someone added the rule for nightly-esr24 so it works). But the > app is querying using aus3.m.o, which is what we want for the release > builds. Bug 1027608 will change this, if we land it there. Yeah...I wasn't planning on backporting that to ESR immediately, but it seems like we should - it's silly to leave these on aus3.m.o until we switch releases over. I'll make sure I do that after it gets review.
The patch for bug 1027608 may not end up helping here. We might have to do something else. I'll take care of this once we know what we're doing over there.
Assignee: nobody → bhearsum
Depends on: 1027608
(In reply to Ben Hearsum [:bhearsum] from comment #8) > The patch for bug 1027608 may not end up helping here. We might have to do > something else. I'll take care of this once we know what we're doing over > there. We've decided to go the route of load balancer redirection rather than in product changes for this. In this case, we'll need to adjust the ifdefs for update server selection only in the esr31 repository. This patch should do it for Firefox.
Attachment #8460427 - Flags: review?(gavin.sharp)
Attachment #8460430 - Flags: review?(standard8)
(In reply to Ben Hearsum [:bhearsum] from comment #9) > We've decided to go the route of load balancer redirection rather than in > product changes for this. In this case, we'll need to adjust the ifdefs for > update server selection only in the esr31 repository. This patch should do > it for Firefox. Why do we need to treat esr-31 differently?
(In reply to :Gavin Sharp [email: gavin@gavinsharp.com] from comment #11) > (In reply to Ben Hearsum [:bhearsum] from comment #9) > > We've decided to go the route of load balancer redirection rather than in > > product changes for this. In this case, we'll need to adjust the ifdefs for > > update server selection only in the esr31 repository. This patch should do > > it for Firefox. > > Why do we need to treat esr-31 differently? My preference is to change update servers in-app wherever possible, because it makes debugging much easier (you know based on client logs exactly what servers you were hitting, rather than depending on knowing what the load balancer configuration at a particular time was). I actually think of this as doing things the same, because in-app is how we've made the switch for nightly/aurora. bug 1027608 is special because it can't be done on the client.
No longer depends on: 1027608
Comment on attachment 8460427 [details] [diff] [review] put nightly esr31 on aus4.mozilla.org You should adjust the comment to explain what's going on here in more detail (as-is it's somewhat misleading). I assume you'll be testing the builds to make sure this is working correctly.
Attachment #8460427 - Flags: review?(gavin.sharp) → review+
(In reply to :Gavin Sharp [email: gavin@gavinsharp.com] from comment #13) > Comment on attachment 8460427 [details] [diff] [review] > put nightly esr31 on aus4.mozilla.org > > You should adjust the comment to explain what's going on here in more detail > (as-is it's somewhat misleading). Will do. > I assume you'll be testing the builds to > make sure this is working correctly. Yeah. I'll do a nightly build and a release build to make sure that they both get the right server, and I'll verify nightlies once this lands.
Attachment #8460430 - Flags: review?(standard8) → review+
So, it turns out that this doesn't work because preprocessor.py doesn't support "-" as part a string to compare. Rather than start down another path I'm just going to leave these updates on aus3 for now, and they can switch when we switch releases over. Gavin, Mark - sorry for your wasted time looking at these patches. Marking this FIXED because the original fix was applied to aus3, and updates are working.
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
Component: General Automation → General
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: