Closed Bug 744729 Opened 12 years ago Closed 11 years ago

Sometimes dashboard update script just hangs, preventing further tests until killed

Categories

(Testing Graveyard :: Eideticker, defect)

All
Linux
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED INVALID

People

(Reporter: wlach, Unassigned)

Details

(Whiteboard: [eideticker:p2])

I haven't diagnosed the exact reason for this, but it is not possible to run further tests if an existing test process has hung. This can cause the dashboard to completely lock up until I manually come in to reset it. I'm not sure exactly why this is: we now time out the actual HDMI capture so as not to fill the disk, so that's not the issue. 

Anyway, regardless of the reason why this is preventing further tests, we just plain shouldn't hang. :) There's a few problems here:

1. Some parts of the various scripts that it calls are susceptible to hangs. For example, we wait indefinitely for the "monkey" script to start here: https://github.com/mozilla/eideticker/blob/master/bin/runtest.py#L79. We should have some kind of timeout. There may be other examples of this as well, need to dig deeper.
2. We really need to just kill everything related to the capture/test after a specified timeout, probably using one of the suggestions here: http://stackoverflow.com/questions/392022/best-way-to-kill-all-child-processes.
Haven't seen this in a while, general phone reliability seems to be a larger issue.
Whiteboard: [eideticker:p2]
Haven't seen this in a year, closing.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → INVALID
Product: Testing → Testing Graveyard
You need to log in before you can comment on or make changes to this bug.