Issue Summary

For large instances, if the backup is running during a deployment or during node termination (due to a number of reasons), it will fail with DEAD status.

This is an intermittent issue that happens with higher frequency for large instances, such as +150GB.

Steps to Reproduce

1. In a Confluence instance with 200GB or more, trigger a backup (including attachments);
2. If the backup completes, trigger it in a different hour of the day.

Expected Results

The backup should always succeed;

Actual Results

The backup fails intermittently with DEAD status and no error logs in the backend.

Workaround

Since this issue happens intermittently, triggering it again in a different time of the day usually works.