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

Issue 735093 link

Starred by 8 users

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: Jun 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 1
Type: ----



Sign in to add a comment

Bugdroid has stopped commenting

Project Member Reported by scottmg@chromium.org, Jun 20 2017

Issue description

There are no comments after commit from bugdroid on monorail issues.

This started sometime over the weekend, maybe 3-4 days ago.
 
Labels: -Restrict-View-Google
Labels: Infra-Troopers

Comment 3 by hinoka@chromium.org, Jun 20 2017

Cc: aga...@chromium.org
Owner: dsansome@chromium.org
Status: Assigned (was: Untriaged)

Comment 4 by hinoka@chromium.org, Jun 20 2017

I see this in the logs:

[E2017-06-20T19:52:56.708760+00:00 26843 140540272154368 infra.services.bugdroid.gitiles_poller:182] Uncaught Exception in <infra.services.bugdroid.bugdroid.BugdroidGitPollerHandler object at 0x7fd25b1e1050>
Traceback (most recent call last):
  File "/b/build/slave/bugdroid/build/infra/infra/services/bugdroid/gitiles_poller.py", line 176, in _ProcessGitLogEntry
    handler.ProcessLogEntry(log_entry)
  File "/b/build/slave/bugdroid/build/infra/infra/services/bugdroid/bugdroid.py", line 102, in ProcessLogEntry
    issue = self.monorail_client.get_issue(project, bug)
AttributeError: 'BugdroidGitPollerHandler' object has no attribute 'monorail_client'


There has been a lot of changes to bugdroid in the last 2 days, I'll do a speculative revert to the state as of last week.

Comment 6 by aga...@chromium.org, Jun 21 2017

The offending changes were reverted here https://chromium-review.googlesource.com/c/540829/

A reland+fix is here https://chromium-review.googlesource.com/c/542277/
Thanks! Any chance of having it comment on the ones it missed or are those lost into the ether?

Comment 8 by aga...@chromium.org, Jun 21 2017

They're lost :(

Bugdroid has two failure modes: failing before polling, or failing after polling. In the former case it can (and will) go back and fill in the gap. Unfortunately this was the latter case.
Status: Fixed (was: Assigned)
I've re-landed the bad changes and added tests to stop this happening again.

Sign in to add a comment