Skip to content

Tech note - 1.68 in-place update failure

When performing an in-place upgrade to any versions between 1.68.13 and 1.68.18 with a very large task history, the upgrade is taking longer than expected due to needing to re-index the task history.

The following workaround can be used to reduce issues when updating.

Archive by truncation

Truncating large tables before an in-place update may allow the in-place update to succeed where otherwise it might have been interrupted by contention.

Truncate the following tables:

  • task_batch_history
  • task_package_history