New issue
Advanced search Search tips

Issue 654836 link

Starred by 1 user

Issue metadata

Status: Available
Owner: ----
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 2
Type: Bug



Sign in to add a comment

Bad Manifest CLs don't get blamed by the CQ.

Project Member Reported by dgarr...@chromium.org, Oct 11 2016

Issue description

Changes to the internal manifest repository are used in the CQ to generate the pinned manifest used for the CQ run.

However, those manifests are applied and "tested" much differently from other CLs, and don't ever get blamed for CQ failures.

They should very clearly be blamed if we are unable to sync code after applying the CL, but that doesn't happen today, as can be seen in the following failures all caused by my CL:*293055. Those failures continued until the CL was manually marked as non-valid.

https://uberchromegw.corp.google.com/i/chromeos/builders/master-paladin/builds/12584
https://uberchromegw.corp.google.com/i/chromeos/builders/master-paladin/builds/12585
https://uberchromegw.corp.google.com/i/chromeos/builders/master-paladin/builds/12586
 
When evaluating priority, the failure mode is to kill the CQ in a really confusing way until there is manual intervention. However, this has come up once in the last 6 months, which is when manifest testing was introduced.

Comment 2 by autumn@chromium.org, Oct 11 2016

Labels: -current-issue
Status: Available (was: Untriaged)
Owner: ----
Components: Infra>Client>ChromeOS>CI
Components: -Infra>Client>ChromeOS

Comment 7 by nxia@chromium.org, Jun 8 2018

Cc: -nxia@chromium.org

Sign in to add a comment