How data flows in continuous replication synchronous mode
In synchronous mode of continuous replication, CVR processes an incoming write by performing the following steps.
CVR receives a write on the primary host.
CVR writes it to the primary SRL.
CVR sends the write to the secondary host and waits for the synchronous network acknowledgments from the secondary hosts. At the same time, CVR writes to the data volumes on the primary host.
On the secondary host, CVR receives the write, processes it, and sends a network acknowledgment to the primary host.
CVR sends writes to the data volumes on the secondary host. When the primary host receives a network acknowledgment from the secondary host, CVR acknowledges to the application that the write is complete.
Note:
The secondary RVG sends the network acknowledgment as soon as the write is received in the CVR kernel memory. This removes the time required to write to the secondary data volumes from the application latency. On the primary host, CVR does not wait for data to be written to the secondary data volumes. This improves application performance. However, CVR tracks all such acknowledged writes that have not been written to the data volumes. CVR can replay these tracked writes if the secondary host crashes before writing to the data volumes on the secondary host or if the primary host crashes before it receives the data acknowledgment.
When the write is written to the data volumes on the secondary host, CVR on the secondary host sends a data acknowledgment to the primary host. CVR marks the write as complete in the SRL when the primary receives the data acknowledgment from all the secondary hosts.