The master is a Win 2003 32 bit installed, while the client is Win 2003 64 bit.
The master and client are at version 6.0MP4. The backup is occuring over primary ip.
There are three local volumes on the client.
ALL_LOCAL_DRIVES is set in backup selection list with multistreaming enabled.
So, we have 3 jobs running for each of the 3 drives and 1 for system state.
However, backup is failing for the drives which hosts the database. Microsoft SQL server 2008 is installed on the client
the error is :
2/22/2011 11:13:46 AM - Error bpbrm(pid=5128) could not write KEEPALIVE to COMM_SOCK
2/22/2011 11:22:22 AM - Error bpbrm(pid=5128) could not write checkpoint processed message to COMM_SOCK
2/22/2011 11:22:23 AM - Error bpbrm(pid=5128) could not send server status message
2/22/2011 11:22:27 AM - end writing; write time: 01:28:34
socket write failed(24) .
Status 24 usually points to network trouble. Review this technote:
also, I would apply MP7 update which fixed a lot of things
If this is an extremely active sql server, you might see that error if you are running a snapshot mode in the backup (if vss or vsp is enabled). The snapshot may be hanging attempting to quiesce but the SQL server is too active for that particular drive. You can disable VSS or VSP snapshotting under the client attributes portion on the master server properties. Just make sure that the client name that you add matches exactly the one in the policy.