Availability Groups SQL Server 2012.

  • Node 1 and Node 2 in primary data centre - sync mode for HA
  • Node 3 in dr data centre - async mode for DR

If a manual failover is performed between node 1 and node 2 can the failover be delayed if there is high "log_send_queue_size" value to node 3?

I would hope not but also don't understand how SQL Server keeps track of the LSNs which still need to be sent when failing over.


When a Secondary (Node 3) is running in Asynchronous-Commit mode, Primary (Node 1) does not wait for a confirmation that Secondary has indeed finished writing the logs to its own log file. It will still wait for a confirmation from a secondary in Synchronous-Commit mode (Node 2) though.

Right after Node 1 will be finished writing the log record to its own log file, it will consider that the transaction has been completed and the client will be notified.

With Asynchronous-Commit mode, Secondary is considered to never be synchronized with the Primary. Secondary always remains unsynchronized and some lag with the Primary is to be expected although it will try to keep up and catch up with the Primary,

Therefore problems or lags on a Secondary in Asynchronous-Commit mode won't impact the Primary.

Much more information are available on this MSDN page: Availability Modes (AlwaysOn Availability Groups)

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.