HADR_RECOVERY_WAIT_FOR_UNDO

Description

For databases participating in an Availability Group. This wait type will accumulate when database recovery is waiting for secondary replicas to become synchronized with the primary replica. This most often occurs after failover (manual or automatic). This wait value will depend on the amount of activity occurring just before failover and how far the secondary was behind at the time. If values are excessive see suggested solutions below

Resolved by

DBAs

Suggested solutions

  1. Ensure secondary replicas have adequate hardware to keep up with the primary
  2. Consider setting the Database Recovery Time to an appropriate value (in line with SLAs)
  3. Check the SQL Server error log for more information

Additional research

Target Recovery Time of a Database
Monitor SQL Server AlwaysOn Availability Groups

See more

Consider using ApexSQL Monitor, a Windows and SQL Server monitoring tool able to track SQL Server wait types, including the individual query wait statistics

⇐ Back to index