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

Issue 607390 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
Closed: Apr 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: iOS
Pri: 1
Type: Bug



Sign in to add a comment

iOS builders are overloaded

Project Member Reported by dpranke@chromium.org, Apr 28 2016

Issue description

This is split off from comments #31 - #33 of bug 407788; it looks like the iOS
bots are overloaded again.

My suspicion is that this is at least partially due to running tests on both
the GN and Ninja bots (GYP) in the CQ. We should only be running tests on the Ninja (GYP)
bots in the CQ, but also run the tests on both waterfall bots.

I'll make that change and then we can check the loads again ...
 
Project Member

Comment 1 by bugdroid1@chromium.org, Apr 28 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/728d1b4e7cf3dcb544a1c6a7dbba204f62d307b9

commit 728d1b4e7cf3dcb544a1c6a7dbba204f62d307b9
Author: dpranke <dpranke@chromium.org>
Date: Thu Apr 28 02:54:46 2016

Stop running tests on the iOS simulator GN trybots.

We don't (currently) have the resources to run the tests
for both the GN and GYP flavors of bots in the CQ, so we
should only run one set of them. Usually we pick the
configuration we currently ship for this, so this CL
turns off the tests on the GN bots.

We will still continue to run the tests on the waterfall bots,
so we should still have coverage. We can (and should) double-
check the load to see how things look after a while to
re-evaluate things.

R=smut@google.com, sdefresne@chromium.org
BUG= 607390 

Review-Url: https://codereview.chromium.org/1932603002
Cr-Commit-Position: refs/heads/master@{#390280}

[modify] https://crrev.com/728d1b4e7cf3dcb544a1c6a7dbba204f62d307b9/ios/build/bots/tryserver.chromium.mac/ios_dbg_simulator_gn.json

Comment 2 by jam@chromium.org, Apr 28 2016

Thanks
Status: Fixed (was: Assigned)
I think we might be okay now. It's a bit hard to say, and it looks like yesterday and the day before were particularly bad for reasons I don't yet know (maybe more patches were failing and so we had to do more retries?).

I'm going to call this fixed for now but hopefully we can check in again next week.

Sign in to add a comment