I didn't set `retry: false` for two reasons:

1. These are enqueued from events as well as cronjobs, so it could
   occasionally help to retry once after 30 seconds.
2. The existing log failure would have to be moved into a rescue if
   `retry: false`, since `sidekiq_retries_exhausted` would not get
   executed by Sidekiq code. And it's not worth the bother.

Changelog: changed
EE: true