Good afternoon. I'm hoping someone can help with a companies query
They have IBM v7000 SAN performing replication via a change log type Lun on the SAN. This goes over a not particularly fast link cross campus to the DR SAN. Replication is set to occur every 5 minutes, however due to bandwidth contention it could well be that it takes 20 minutes + to transmit 5 minutes worth of changes stored on the primary San. The question is, how does VMware handle the VMDK files at the DR site in the event of a primary site failure if not all the recent changes have replicated over? Customer is concerned that a primary site failure would corrupt the DR site VMDK's if a failure occurred during one of these slow / delay replication slots.
Does VMware maintain the integrity of the DR VMDK files etc until ALL expected changes are successfully received? Essentially a last known good copy?