1534055 - crontabber needs access to S3 crash bucket

Status: RESOLVED FIXED

(Socorro :: Infra, task, P2)

Description

[Will Kahn-Greene [:willkg] ET needinfo? me]

In bug #1528243, I wrote a crontabber app that lists the keys for raw crash files for a specified day in the S3 crash bucket and then figures out which ones weren't processed.

I pushed that out today and it kicked this up:

https://sentry.prod.mozaws.net/operations/socorro-new-stage/issues/5337472/

This bug covers fixing the crontabber node in stage and prod to have list access to the S3 crash bucket.

Comment 1

[Will Kahn-Greene [:willkg] ET needinfo? me]

The code needed s3:GetObject access as well as s3:ListBucket access.

After that, it ran fine:

2019-03-11T19:42:18.854136 INFO socorro.cron.base.VerifyProcessedCronApp: All crashes for 20190310 were processed.

It's definitely running now.

I spun off bug #1534402 to fix the code to wrap the multiprocessing code in Sentry error handling.

Comment 2

[Will Kahn-Greene [:willkg] ET needinfo? me]

Backfilling the priority.

Comment 3

[Will Kahn-Greene [:willkg] ET needinfo? me]

I fixed error handling (I think). I think we're good here now.