New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 639318 link

Starred by 1 user
Status: Fixed
Owner:
charliea@chromium.org
Closed: Aug 2016
Cc:
dtu@chromium.org
charliea@google.com
█ aschulman@chromium.org
mac-bugs-priority@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: Mac
Pri: 1
Type: Bug



Sign in to add a comment

BattOr power data is not being collected for Mac Retina tryserver (build28-b4)

Project Member Reported by charliea@chromium.org, Aug 19 2016 
Link to try job: https://uberchromegw.corp.google.com/i/tryserver.chromium.perf/builders/mac_retina_perf_bisect/builds/1588

Note that I launched this try job at a time when BattOr perf tests were disabled on the main waterfall, thus my use of --also-run-disabled-tests. Without this flag, I'd expect my test wouldn't have run because a BattOr was not detected.

The reason why I think that the BattOr didn't run was because, in the traces collected (e.g. https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/trace-file-id_11-2016-08-18_14-17-34-24942.html), there seems to be no power data.

Strangely, whatever this problem is seems to be isolated to the tryservers: the main perf waterfall servers are humming along nicely and collecting power data.

Comment 1 by sullivan@chromium.org, Aug 19 2016

Cc: dtu@chromium.org
Adding dtu in case he has any clue what debugging steps to try.

Aaron, is there some easy logging we could add to debug? Since it's an unsubmitted CL, we don't need to put log statements through code review, we can just try a ton of things and fire away to the trybots.

Comment 2 by charliea@chromium.org, Aug 19 2016 

I requested MDB access to the try-build-access group earlier today, and pschmidt@ just approved it. I'm going to try to SSH into the bot and see if I can see the device listed among the serial devices.

Comment 3 by charliea@chromium.org, Aug 19 2016

Components: Infra>Labs
After a little investigating, I don't think the BattOr is connected via USB to the host. When I SSH into the buildbot and do:

$ ls /dev/tty*usbserial*

I get: 

ls: /dev/tty*usbserial*: No such file or directory

Compare this to build27-b4 (or build29-b4):

$ ls /dev/tty*usbserial*
/dev/tty.usbserial-MAAB-AAA4

Adding Infra-labs. Could someone from Infra-labs make sure that the USB cable from the BattOr to the host is plugged in? If it is, could you try swapping out another USB cable? We've had problems with some in the past.

Comment 4 by pschm...@google.com, Aug 19 2016 

usb cable wasn't seated in the laptop's usb port.

chrome-bot@build28-b4:(Mac 10.11.6):~$ ls /dev/tty.usbserial*
/dev/tty.usbserial-MAAB-AAAJ

Comment 5 by charliea@chromium.org, Aug 19 2016 

Thanks Peter! You're a lifesaver.

Comment 6 by charliea@chromium.org, Aug 19 2016

Status: Fixed (was: Untriaged)

Comment 7 by charliea@chromium.org, Aug 19 2016

Owner: charliea@chromium.org
►

Sign in to add a comment