At least one secondary database replica is lagging behind its primary replica. The alarm identifies the affected database, the secondary replica, and the estimated synchronization delay, in seconds, relative to the primary replica.
This alarm indicates that an Availability Group secondary replica is lagging behind its primary replica. The alarm estimates the time required to transmit pending transaction log data to the secondary replica and apply (redo) the changes. Latency is calculated based on the current sizes and processing rates of the log-send and redo queues.
Severity thresholds
- Normal: Less than 60 seconds behind
- Medium: 60 to less than 300 seconds behind
- High: 300 seconds or more behind
High latency may indicate slow network throughput, insufficient disk or CPU resources on the secondary replica, or a synchronization process that is stalled or progressing slowly.
Note: Synchronization latency is calculated only when SQL Server reports a log-send or redo rate greater than zero. When the reported rate is zero, synchronization latency is considered zero and no alert is triggered. As a result, an actual synchronization stall may not trigger an alert because it cannot be distinguished from normal inactivity.