New issue
Advanced search Search tips

Issue 721890 link

Starred by 1 user

Issue metadata

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



Sign in to add a comment

Consider disabling CQ auto-retries in the event of failure due to capacity constraints

Project Member Reported by jbudorick@chromium.org, May 12 2017

Issue description

Immediately retrying a build that failed due to a capacity problem is likely to exacerbate said problem.
 
Components: -Infra Infra>CQ
(not sure about component here)
Interesting idea. Before I do this, I'd want to see these stats:

% of CQ attempts on which this retry happened
% of CQ attempts on which this retry happened AND attempt succeeded 


btw, John, don't mark bugs available or else I wouldn't notice them!
Labels: -Type-Task Type-Feature
ah, oops. Untriaged next time.
Components: -Infra>CQ Infra>Platform>CQdaemon

Comment 6 by efoo@chromium.org, Aug 31 2017

Components: Infra>Platform>CQ

Comment 7 by efoo@chromium.org, Aug 31 2017

Components: -Infra>Platform>CQdaemon

Comment 8 by efoo@chromium.org, Jun 2 2018

Friendly ping. This is a blocking bug for cit-pm-43. 
This isn't a trivial feature -- it's not clear how CQ can reliably learn of capacity issues. Nor do we have broad agreement that this has to be done. So, design doc is in order.
ah, but I don't have time in Q2 given high Pri1 for other things even in CQ.

Sign in to add a comment