Replication latency for secondary servers

When you combine Enterprise Replication with high-availability clusters, replication latency can increase.

When Enterprise Replication is running on a high-availability cluster, some operations cannot be performed until the logs are shipped to the secondary server. By default, the logs are shipped to secondary servers after 50 replicated transactions are applied, or 5 seconds elapse. This delay prevents possible inconsistency within the Enterprise Replication domain during a failover to a secondary server.

You can control replication latency for high-availability data replication (HDR) servers in one of the following ways
  • Set HDR replication to fully synchronous, nearly synchronous, or asynchronous mode.
  • Set HDR replication to HDR SYNC.
  • Adjust the DRINTERVAL configuration parameter to specify a different interval between flushing the high-availability data-replication buffer.

If you combine Enterprise Replication with shared-disk secondary servers, you can reduce replication latency by setting the CDR_MAX_FLUSH_SIZE configuration parameter to 1 to flush the logs after each replicated transaction.


Copyright© 2019 HCL Technologies Limited