Closed Bug 937672 Opened 11 years ago Closed 11 years ago

unable to run update tests through a testing channel

Categories

(Mozilla QA Graveyard :: Mozmill Automation, defect)

x86
macOS
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED INVALID

People

(Reporter: jbecerra, Unassigned)

References

Details

For a few days I have been trying to run a type of update test scenario, but I haven't been able to get the automation to run to completion.

The config file for this update is as follows:

[testrun]
script=update
report=http://mozauto.iriscouch.com/mozmill-release/
target-build-id=20131021192059
channel=brbtest

[mac 10.6 64bit]
platform=mac
25.0b12=en-US de es-ES

It is supposed to install Firefox 25 beta 12 and try to update through "brbtest" and the resulting build should have the build id above which corresponds to the first candidate builds for Firefox 25. That is all I would like to check for now, as part of testing bug 908134.

Unfortunately when I go through the usual procedure I use to test, say, updates for a beta release through betatest, for example, I am not able to get results using this config file.

The console output is as follows:

08:16:17 Started by user anonymous
08:16:17 [EnvInject] - Loading node environment variables.
08:16:17 Building on master in workspace /data/mozmill-ci/jenkins-master/jobs/trigger-ondemand/workspace
08:16:17 Copying file to ondemand.cfg
08:16:17 [workspace] $ /bin/sh -xe /tmp/hudson7229358802168353215.sh
08:16:17 + /data/mozmill-ci/./jenkins-master/../jenkins-env/bin/python trigger.py ondemand.cfg
08:16:17 Triggering job: ondemand_update with {'PLATFORM': 'mac', 'REPORT_URL': 'http://mozauto.iriscouch.com/mozmill-release/', 'VERSION': '25.0b12', 'BUILD_NUMBER': '1', 'ENV_PLATFORM': 'mac', 'LOCALE': 'en-US', 'TARGET_BUILD_ID': '20131021192059', 'NODES': 'mac && 10.6 && 64bit', 'BUILD_TYPE': 'release', 'CHANNEL': 'brbtest'}
08:16:17 Traceback (most recent call last):
08:16:17   File "trigger.py", line 98, in <module>
08:16:17     main()
08:16:17   File "trigger.py", line 95, in main
08:16:17     j.build_job('ondemand_%s' % script, parameters)
08:16:17   File "/data/mozmill-ci/jenkins-env/local/lib/python2.7/site-packages/jenkins/__init__.py", line 456, in build_job
08:16:17     self.build_job_url(name, parameters, token)))
08:16:17   File "/data/mozmill-ci/jenkins-env/local/lib/python2.7/site-packages/jenkins/__init__.py", line 218, in jenkins_open
08:16:17     'Possibly authentication failed [%s]' % (e.code)
08:16:17 jenkins.JenkinsException: Error in request.Possibly authentication failed [500]
08:16:17 Build step 'Execute shell' marked build as failure
08:16:17 Finished: FAILURE

Should I be able to run this test through the automation and get a result?
As discussed on IRC this is because the ondemand_update job only accepts a defined set of values for the update channel. These are:

nightly
aurora
auroratest
beta
betatest
release
releasetest
esr
esrtest

We can either add more to this list, or make this field accept any value. Please raise an issue here https://github.com/mozilla/mozmill-ci/issues to address this.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → INVALID
Nothing in the failure above indicates that it is a failure in the update script. There is a Jenkins authentication issue listed. So we are already failing to submit the job to Jenkins. But it's true, this should have been filed in mozmill-ci.
The change is active on staging now and will be pushed to production sometime late this week or early next week. Juan, do you still have to run those update tests, or not anymore for now?
Not for now, but I think a similar scenario will have to be tested in the coming weeks. Thank you!
Product: Mozilla QA → Mozilla QA Graveyard
You need to log in before you can comment on or make changes to this bug.