Re: Restarting tryserver.chromium.linux

[Viet-Trung Luu]

-> the correct chromium-dev (@chromium.org)

On Mon, Aug 18, 2014 at 1:47 PM, Viet-Trung Luu <viettr...@chromium.org> wrote:

On Mon, Aug 18, 2014 at 1:45 PM, Aneesh Mulye <ane...@google.com> wrote:

I'd like to restart tryserver.chromium.linux. This is to pick up the
addition of a linux_chromium_browser_asan_rel recipe.

This e-mail is to:
a) Notify users who may have jobs in the run-queue (the instructions
for finding out who specifically has a job there doesn't seem to be
working), and
b) To request sherrif sign-off.

Could one of the sherrifs authorise this?

SGTM

[John Abd-El-Malek]

I think we should wait till we have linux_chromium_asan_rel_swarming (i.e. one bot that builds and runs all of the current linux tests through swarming). https://code.google.com/p/chromium/issues/detail?id=404694

[Chase Phillips]

We typically do not wait to restart for other configs that aren't ready now (CL posted, reviewed, has lgtm) since restarting again later in the evening or the next day for those is low cost.  Is there a lgtm'd CL for the asan bot now?

If not, we shouldn't block the tryserver.chromium.linux restart on the linux_chromium_asan_rel_swarming change.  If that asan bot CL is ready by the time of Aneesh's restart, then we'll land it and pick it up at the same time as this restart.  Otherwise tryserver.chromium.linux will just need another restart later for the asan bot CL, and that's fine.

[John Abd-El-Malek]

What I had meant was that there's not much point in having the non-swarming browser-tests asan linux config. The existing non-recipes can suffice until we have a config that we can get useful timing measurements for in terms of adding it to the CQ.