New issue
Advanced search Search tips

Issue 698087 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
Closed: Mar 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 2
Type: Bug



Sign in to add a comment

Fix incorrect assumption about |machine| argument to server_job

Project Member Reported by pprabhu@chromium.org, Mar 3 2017

Issue description

See  issue 685870  for an instance of this problem that made me look for others.

At some point (a year ago), we changed the |machine| argument to server_job from a string containing the hostname of the host to possibly a dict containing more stuff.

Some server tests used |machine| as the hostname but were not updated.

Here's a quick list of broken control files from some grepping:
(all paths relative to server/site_tests/)

platform_USBHIDWake/control
kernel_MemoryRamoop/control
bluetooth_RegressionServer/control
network_DiskFull/control
power_BatteryStateOnResume/control
hardware_StorageQualTrimStress/control.test
hardware_StorageQualSuspendStress/control.test
network_EthCapsServer/control
hardware_StorageStress/control (and other control files for that test)
hardware_StorageQualBase/control.test
hardware_MemoryIntegrity/control (and other control files for that test)
sequences/control.memory_qual ( issue 685870 )
sequences/control.power_build
sequences/control.power_weekly
sequences/control.memory_qual_quick
sequences/control.power
sequences/control.power_daily
stress_ClientTestReboot/control.touchUpdate
stress_ClientTestReboot/control.bluetooth

Here's an example of where this _was_ fixed correctly:
http://shortn/_cldzZroHyv

 
I'm a bit surprised by the larger number of tests on that list.

+sbasi: can you spot check a few of these and verify that my reading of the control files is correct.
The problem is that they use |machine| and call some function that expects a hostname. This may not fail in _all_ situations (iirc, in some cases |machine| is still just a hostname) but will fail when run via, say suite_scheduler.
Cc: dbasehore@chromium.org
+dbasehore: You know the fix that I applied for  issue 685870 .

I do not intend to take any action on this bug at the present. If there are other tests that you care about that are failing in a similar fashion, please apply the same fix. Feel free to send the CL to me for review.

Comment 3 by autumn@chromium.org, Mar 15 2017

Labels: -current-issue
Owner: dbasehore@chromium.org
Owner: ----

Comment 5 by autumn@chromium.org, Mar 21 2017

Status: Fixed (was: Available)
Solution has been offered, marking as fixed. 

Comment 6 by autumn@chromium.org, Mar 21 2017

Owner: norvez@chromium.org
Status: Available (was: Fixed)
Revised: Passing to sheriff to follow up if an owner is needed

Comment 7 by norvez@chromium.org, May 17 2017

Cc: norvez@chromium.org
Owner: dbasehore@chromium.org
Status: Assigned (was: Available)
Re-assigning to dbasehore@.
Derek, can you close the bug if the solution in #c2 is OK?
Status: Fixed (was: Assigned)
presumed fixed

Sign in to add a comment