Aws::S3::Errors::NoSuchKey

What happened

• The job lists every file under elevance-tn/unprocessed, then reads each one with get_blob.
• S3 returned NoSuchKey for a file that had been listed a moment earlier, so it was probably moved or deleted in between.
• Only 13 events in 110 days on a job that runs hourly means this is a rare race, not a permanent bug.
• The job's rescue block writes an ErrorLog and marks the audit Failed, then re-raises. The next hourly run lists only the files still unprocessed.

What to do next

1. Query Shows whether runs overlapped, and whether the next run after the failure finished successfully.

   SELECT id, start_time, end_time, comment, audit_data FROM execution_audits WHERE auditable_type = 'WellpointTnCensusJob' AND start_time > now() - interval '3 days' ORDER BY start_time DESC;

2. Check Check Bugsnag and the Sidekiq logs for the failing runs. Look at whether two WellpointTnCensusJob runs started close together, or whether a deploy or restart happened at that time (the last event was at 00:03 UTC on 2026-11-11, near release 20261110135930).

Queries are shown, never run. Check them before using them on production.