ReleaseEngineering/Buildduty/SVMeetings/Oct5-Oct9

From MozillaWiki
Jump to: navigation, search

Upcoming vacation/PTO: Canadian Thanksgiving (holiday): Oct 12 coop - Oct 13-16 vlad - oct16 - oct20 alin - dec 24, dec 28-31 florin - Oct 15-16 otilia - Oct 9

Meetings every Tuesday and Thursday

https://wiki.mozilla.org/ReleaseEngineering/Buildduty/SVMeetings/Sept28-Oct2 https://wiki.mozilla.org/ReleaseEngineering/Buildduty/SVMeetings/Sept28-Oct2-coop-visit


2015-10-05 - 2015-10-06

[vlad] slave api https://wiki.mozilla.org/ReleaseEngineering/Applications/SlaveAPI https://api.pub.build.mozilla.org/docs/usage/slaveloan/ https://github.com/mozilla/build-relengapi

https://bugzilla.mozilla.org/show_bug.cgi?id=1158729 

           Q: Kim can i close the bug ticket ?  
           Update: I spoke with Kim on irc and she confirmed that everything is ok. The bug has been closed

https://bugzilla.mozilla.org/show_bug.cgi?id=1203128 the patches have been created and attached to the bug yosemite-r7 slaves added to slavealloc db will land changes tomorrow morning and reconfig. Then you can run some pushes on try to look at test results modify check_pending_builds to report more granularly on pending builds/tests ( https://bugzilla.mozilla.org/show_bug.cgi?id=1204970

.

[alin] https://bugzilla.mozilla.org/show_bug.cgi?id=1211808 - loan request for a panda slave are we able to loan such slaves? tried to connect to several panda slaves, but some of them are refusing the connection right away and in other cases the connection credentials are not working (tried even with kickstart) 

     2.   when loaning a slave, we send an e-mail to the developer in which we suggest the following link for VPN setup:

https://mana.mozilla.org/wiki/pages/viewpage.action?pageId=30769829 this link would need some adjustments, since we switched to MFA authentication at the moment should we do the change/is there somebody else we would need to talk to? open a bug against DCOPs to update the bug 

     3. https://bugzilla.mozilla.org/show_bug.cgi?id=1210822

work in progress, trying to understand the process of adding graphene builds from bug 1130090

Other bugs to work on Could you look at this bug please https://bugzilla.mozilla.org/show_bug.cgi?id=1210822 (remove larch graphene build)

I ran another try build tonight for the 10.10.5 tests. The one earlier today did not invoke tests. https://treeherder.mozilla.org/#/jobs?repo=try&revision=0084ba8c2789

I wrote a short two line thing here on how to loan pandas - not sure if you still need this https://wiki.mozilla.org/ReleaseEngineering/How_To/Android_Tegras#To_loan_pandas


2015-10-07 - 2015-10-08

[alin] https://bugzilla.mozilla.org/show_bug.cgi?id=1210822 - Remove larch graphene builds 

   - looked over bug 1130090 and 1160717 to understand the changes that were made
   - what we want to do is to stop running these builders on larch but keep them running on mozilla-central
   - looked in treeherder on larch: https://treeherder.mozilla.org/#/jobs?repo=larch and saw several changesets that included graphene builders, but these changesets do not occur too often (the last one is from Sep 18)
   - did the same thing on mozilla-central: https://treeherder.mozilla.org/#/jobs?repo=mozilla-central -> scrolled down up until around ~August 26, did not see any graphene build, so I don't know if this is expected or not
   - my idea was to analyze the config files of several buildbot masters that have graphene builders attached and after that write a patch to disable those builders on larch, but I don't seem to get the things right
   - I created a new test master (/builds/buildbot/alin.selagea/graphene_test) and edited "config.py" and "master_localconfig.py" to limit the pool of builders generated after starting the master to only one branch ("larch"). Unfortunately, I did not have much luck on this.

Look at http://hg.mozilla.org/build/buildbot-configs/file/6b8028f5a2bf/mozilla/b2g_project_branches.py#l155 try removing the graphene definitions here and comparing the builders before and after UPDATES: tried the above solution as well, there is no difference between the builders generated using the default configs and after removing graphene definitions the problem is that the default configuration (the one that we get when creating the test master) does not generate any larch builders compared different config files from <path_to_master>/master and noticed that some of them are different (even though the links were created several moths ago, they point to files from buildbot-configs repo, which are up-to-date)

Vlad:

The second try run had similar test failures to the one from last night see https://treeherder.mozilla.org/#/jobs?repo=try&author=kmoir@mozilla.com and compare the last two runs

The compare results from both runs: Failing jobs: -> mozilla-central opt test gtest -> mozilla-central debug test gtest Warning jobs: -> mozilla-central opt test mochitest-devtools-chrome-2 -> mozilla-central opt test mochitest-gl -> mozilla-central opt test web-platform-tests-2 -> mozilla-central opt test web-platform-tests-3 -> mozilla-central opt test web-platform-tests-4 -> mozilla-central opt test web-platform-tests-reftests -> mozilla-central debug test mochitest-2 -> mozilla-central debug test mochitest-devtools-chrome-3 -> mozilla-central debug test mochitest-gl -> mozilla-central debug test web-platform-tests-2 -> mozilla-central debug test web-platform-tests-3

update bug of test failures, ask in #ateam for what developers could help look at test failures

Opened bugs so you could have level 1 commit access to run tests on try https://bugzilla.mozilla.org/show_bug.cgi?id=1212493 (Vlad) attached the public key https://bugzilla.mozilla.org/show_bug.cgi?id=1212494 (Alin) also attached the public key

Slave Loan Tool: https://github.com/mozilla/build-relengapi/issues/351

Retrieved from "https://wiki.mozilla.org/index.php?title=ReleaseEngineering/Buildduty/SVMeetings/Oct5-Oct9&oldid=1100123"