Closed
Bug 753260
Opened 12 years ago
Closed 12 years ago
checkStalled/stopStalled not cleaning up everything it should
Categories
(Release Engineering :: General, defect)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: Callek, Assigned: Callek)
References
Details
Attachments
(3 files, 1 obsolete file)
1.43 KB,
patch
|
bear
:
review+
Callek
:
checked-in+
|
Details | Diff | Splinter Review |
1.22 KB,
patch
|
bear
:
review+
Callek
:
checked-in+
|
Details | Diff | Splinter Review |
3.19 KB,
patch
|
bear
:
review+
armenzg
:
feedback+
Callek
:
checked-in+
|
Details | Diff | Splinter Review |
So, the sut_tools checkStalled/stopStalled doesn't clean everything it should. It currently only notices bcontroller.py lines where we use the SUT IP itself. Also the functions require a SUT_NAME in all cases, and in cleanup we pass an IP instead when run from buildbot. I'll address both issues in this bug.
Assignee | ||
Comment 1•12 years ago
|
||
Attachment #622316 -
Flags: review?(bear)
Assignee | ||
Comment 2•12 years ago
|
||
So, since we don't have sut_name available as a build property I decided to cheat and do slavename since we currently identify it as tegra-XXX to match. I can instead, modify the job to set a property, or modify cleanup to take no args and auto-use os.env['SUT_NAME']. If you'd like.
Attachment #622318 -
Flags: review?(bear)
Updated•12 years ago
|
Attachment #622316 -
Flags: review?(bear) → review+
Comment 3•12 years ago
|
||
Comment on attachment 622318 [details] [diff] [review] [buildbotcustom] use tegra name this information is found in two environment variables set when calling each buildbot: SUT_NAME and SUT_IP we may not even need to set this property
Attachment #622318 -
Flags: review?(bear) → review-
Assignee | ||
Comment 4•12 years ago
|
||
This stops us from passing anything to cleanup.py, will require a change in tools cleanup.py though. Deployed the tools repo change won't require tegra downtime, so not a huge rush, I will get it pushed/reviewed before I check this in though.
Attachment #622318 -
Attachment is obsolete: true
Attachment #622420 -
Flags: review?(bear)
Comment 5•12 years ago
|
||
Comment on attachment 622420 [details] [diff] [review] [buildbotcustom] v2 - Pass nothing to cleanup.py doesn't stop us from passing something in the future, just wanted to make all of the sut helper scripts use the same method for knowing what tegra to use. thanks
Attachment #622420 -
Flags: review?(bear) → review+
Assignee | ||
Comment 6•12 years ago
|
||
Comment on attachment 622316 [details] [diff] [review] [tools] v1 this one landed and deployed wednesday
Attachment #622316 -
Flags: checked-in+
Assignee | ||
Updated•12 years ago
|
Attachment #622316 -
Attachment is patch: true
Assignee | ||
Comment 7•12 years ago
|
||
I decided to also do the updateSUT.py adjustments while I was here, since it just made sense to do them both. contexually this patch will conflict with the sys.stdout.flush() patch, but I'll properly rebase when that lands before I land this. It has been tested in staging with various forms of calling
Attachment #623212 -
Flags: review?(bear)
Attachment #623212 -
Flags: feedback?(armenzg)
Updated•12 years ago
|
Attachment #623212 -
Flags: review?(bear) → review+
Comment 8•12 years ago
|
||
Comment on attachment 623212 [details] [diff] [review] [tools] Part 2 -- make <sut_name> arg optional Review of attachment 623212 [details] [diff] [review]: ----------------------------------------------------------------- Thanks a lot for making this saner. ::: sut_tools/updateSUT.py @@ +14,5 @@ > target_version = "1.07" > apkfilename = "sutAgentAndroid.apk" > +tegra_name = os.getenv('SUT_NAME') > +apkFoopyDirPattern = "/builds/%(tegra_name)s" > +apkFoopyDir = apkFoopyDirPattern % {'tegra_name': tegra_name} It's weird to have these 3 variables under #Constants when in main can be modified but I guess it is like this because of the case when we import updateSUT.py rather than calling it directly. @@ +77,5 @@ > print "INFO: updateSUT.py: We're now running %s" % ver > return RETCODE_SUCCESS > > +def main(device): > + dm = connect(device) I would personally call it "tegra_name" instead of "device" since "device" sounds more generic.
Attachment #623212 -
Flags: feedback?(armenzg) → feedback+
Assignee | ||
Comment 9•12 years ago
|
||
Comment on attachment 623212 [details] [diff] [review] [tools] Part 2 -- make <sut_name> arg optional Armen, fyi I'm removing your second hunk in http://hg.mozilla.org/build/tools/rev/16fc4f354b44 So that things work "right" here.
Assignee | ||
Comment 10•12 years ago
|
||
Comment on attachment 623212 [details] [diff] [review] [tools] Part 2 -- make <sut_name> arg optional http://hg.mozilla.org/build/tools/rev/c5e085fcde40
Attachment #623212 -
Flags: checked-in+
Assignee | ||
Comment 11•12 years ago
|
||
Comment on attachment 622420 [details] [diff] [review] [buildbotcustom] v2 - Pass nothing to cleanup.py argh this dropped off my list of checkins last week, this bug can close when it gets merged/pushed to prod. http://hg.mozilla.org/build/buildbotcustom/rev/75dbae83f458
Attachment #622420 -
Flags: checked-in+
Comment 12•12 years ago
|
||
This went went into production today.
Status: NEW → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
Updated•11 years ago
|
Product: mozilla.org → Release Engineering
Updated•6 years ago
|
Component: General Automation → General
You need to log in
before you can comment on or make changes to this bug.
Description
•