When an input journal file gets corrupted, Journalbeat logs are filled with the following message:

Error while reading event: failed to iterate journal: bad message {"id": "{{ uuid }}"}

As journals does "rotate-on-corruption", we can reopen the journal file and Journalbeat can continue reading correct journal entries.

The current workaround is to restart Journalbeat on this error manually or restart the Beat periodically to make sure it can restart after it reaches a corrupted entry.