Most GPU FYI bots purple with slave lost exception |
||||
Issue descriptionMost GPU FYI bots are purple with a slave lost exception. So far this is one build on each bot, and slave lost exception is on different steps on each bot. For example: https://build.chromium.org/p/chromium.gpu.fyi/builders/Mac%2010.10%20Release%20%28AMD%29/builds/633 https://build.chromium.org/p/chromium.gpu.fyi/builders/Mac%2010.10%20Release%20%28Intel%29/builds/21050 https://build.chromium.org/p/chromium.gpu.fyi/builders/Win7%20x64%20Debug%20%28NVIDIA%29/builds/6723 Was there a one-time outage that caused all the bots to go down? Waiting to see if the bots turn green on the next build.
,
Jan 26 2017
No, slave lost is because the connection between master and slave was interrupted, possibly by master restart: https://build.chromium.org/p/chromium.gpu.fyi/builders/Mac%2010.10%20Release%20%28AMD%29/builds/633/steps/steps/logs/stdio remoteFailed: [Failure instance: Traceback (failure with no frames): <class 'twisted.internet.error.ConnectionLost'>: Connection to the other side was lost in a non-clean fashion.
,
Jan 26 2017
Bots are cycling green, looks like the was a one-time purple per bot. A master restart would explain it.
,
Jan 27 2017
Sorry about that. I restarted the waterfall early today (Pacific time, though probably not early for the East Coast troopers) to deploy a new bot in http:///crbug.com/680754 . Should have notified the pixel wrangler. Will do so on the bug next time. |
||||
►
Sign in to add a comment |
||||
Comment 1 by ajuma@chromium.org
, Jan 26 2017