Using a replication between four active servers we find that the LSN numbers on some of the servers are signfiicantly different than others.
None of the documentation points me to any way which shows whether that is a normal, or acceptable state to be.
Replication appears to be broken, i.e. we insert data into one of the DBs and it gets to two of other databases, but not the third. The third database appears to be so far out of replication we need to re-instantiate it, but there is no replication log growth, and no indication anywhere from TimesTen that there is a problem.
The only clue we have that there is something wrong is, the obvious major data discrepancy in one of the four servers, but also that the LSN numbers are so different,
Server 1
Process ID : 9761 (Started)
Replication Agent Policy : always
Host : HOST1
RepListener Port : 43276 (AUTO)
Last write LSN : 41337.39813384
Last LSN forced to disk : 41337.39813120
Replication hold LSN : 41337.38906120
Server 2
Process ID : 32109 (Started)
Replication Agent Policy : always
Host : HOST2
RepListener Port : 34386 (AUTO)
Last write LSN : 56143.198199560
Last LSN forced to disk : 56143.198043648
Replication hold LSN : 56143.184765840
Are the LSNs unique to the server? Are the numbers not related to each other across servers?