Use better delay and retry count for releasetasks submission to taskcluster

RESOLVED FIXED

Status

defect
RESOLVED FIXED
3 years ago
3 years ago

People

(Reporter: Callek, Assigned: Callek)

Tracking

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(1 attachment)

Assignee

Description

3 years ago
We have been getting ISE 500's from the "old" taskcluster scheduler lately trying to submit our releases.

This is likely due to the codepaths being barely used in general now on TC's side.

After speaking with jonas in #taskcluster the submission is idempotent given same inputs, and the jobs do eventually kick off, so with some better backoff/retry logic we can avoid a failed release due to this issue.

(The code is substantially faster validating that a task was already created with a given ID than it is with actually creating said tasks).
Comment hidden (mozreview-request)

Comment 2

3 years ago
mozreview-review
Comment on attachment 8780110 [details]
Bug 1293744 -- fix typo from .update() returning None rather than itself.

https://reviewboard.mozilla.org/r/70910/#review68336
Attachment #8780110 - Flags: review?(rail) → review+
Assignee

Comment 3

3 years ago
https://hg.mozilla.org/build/tools/rev/8686d725cd8c
Status: NEW → RESOLVED
Closed: 3 years ago
Resolution: --- → FIXED
Comment hidden (mozreview-request)

Comment 5

3 years ago
mozreview-review
Comment on attachment 8780110 [details]
Bug 1293744 -- fix typo from .update() returning None rather than itself.

https://reviewboard.mozilla.org/r/70910/#review68450
You need to log in before you can comment on or make changes to this bug.