New issue
Advanced search Search tips

Issue 877097 link

Starred by 1 user

Issue metadata

Status: Duplicate
Merged: issue 876579
Owner: ----
Closed: Aug 23
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 1
Type: ----



Sign in to add a comment

Will -provision-labels provision a DUT with a cros-version label?

Project Member Reported by xixuan@chromium.org, Aug 23

Issue description

I checked bot:

https://chrome-swarming.appspot.com/bot?id=chromeos-skylab-bot-cafaa769-50cb-4a7f-99c3-12dd99fd7f9a&sort_stats=total%3Adesc

As we use fallback request for CQ jobs now, this bot first fails a test, then get repaired with reef-release/R68-10718.71.2.

Then it starts to accept CQ jobs, and run it in the second slice due to it doesn't have correct paladin cros-version:

https://chrome-swarming.appspot.com/task?id=3f7ed8e0c54b9310&refresh=10&show_raw=1

and it passes.

Then it accept another CQ job, and now it should already have the right cros-version "reef-paladin/R70-10993.0.0-rc4", so the next CQ job should run in the first slice? But it runs still in the second slice with provision:

https://chrome-swarming.appspot.com/task?id=3f7eea3c650d3810&refresh=10&show_raw=1

Is it expected?
 
hostinfo sync works normally.
The hostinfo from provision jobs also gets written correctly
Maybe a swarming bug?  See https://chrome-swarming.appspot.com/task?id=3f80ac8238164210&refresh=10

The bot dimensions do have provisionable-cros-version: reef-paladin/R70-10994.0.0-rc2

Not sure why the first slice doesn't run.  Maybe needs a longer timeout on the first slice?
Mergedinto: 876579
Status: Duplicate (was: Untriaged)
Re #4, let me change the timeout distribution of the 2 slices and give it a try.

Sign in to add a comment