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

Issue 721210 link

Starred by 1 user

Issue metadata

Status: Archived
Owner: ----
Closed: Mar 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 2
Type: Bug



Sign in to add a comment

CleanUpStage took 2 hours on butterfly-paladin

Project Member Reported by akes...@chromium.org, May 11 2017

Issue description

Stage: https://uberchromegw.corp.google.com/i/chromeos/builders/butterfly-paladin/builds/22295/steps/CleanUp/logs/stdio

Build details: https://viceroy.corp.google.com/chromeos/build_details?build_id=1508466

Build: https://uberchromegw.corp.google.com/i/chromeos/builders/butterfly-paladin/builds/22295


Snippet:

17:53:56: INFO: Cleaning up buildroot git repositories.
17:53:56: INFO: Wiping old output.
17:53:56: INFO: Deleting archived trybot images.
17:53:56: INFO: Deleting archived perf results.
17:53:56: INFO: Deleting autotest site packages.
17:53:56: INFO: Deleting chroot.
17:55:09: INFO: Running cidb query on pid 30182, repr(query) starts with <sqlalchemy.sql.expression.Update object at 0x7fb731907d10>
19:55:20: INFO: RETRYING cidb query due to (OperationalError) (2026, 'SSL connection error: socket layer receive error') None None.


The 2 hours seems to have been spent trying to talk to cidb? That's very surprising if so. Maybe the timeout / retry numbers on it have been adjusted, but it used to retry a lot more often then that.

This CQ run passed, but this build ended up being the long pole by a lot.
 
Looks more like the call took 2 hours to return a socket error back from cidb on the original request. Very weird.

We don't have timeouts on cidb calls from the client side, and I haven't seen this before.

Comment 2 by aut...@google.com, Jun 2 2017

Labels: -current-issue
Status: Unconfirmed (was: Untriaged)
Waiting for this to recur.
Status: Archived (was: Unconfirmed)
Bulk closing old unconfirmed issues.

Sign in to add a comment