The customer has noticed that some items going to one child server are not replicating during differential replications. All other child servers work fine and most
items work correctly to the one server.
They have noticed the following:
There were some network issues between the parent SMP and the child SMP. The replication process kept retrying to connect causing a backlog of replicated items.
ITMS 8.1 and later
In order to improve the replication process, we had added a new Core setting (ReplicationWebServiceRetryCount) which indicates how many retries should be performed during replication start in case if destination server is not available.
The default value is 10 with 10 seconds delay (delay is incremented after each retry).
With this enhancement, the replication process will retry to replicate by default up to 10 times and fail up after that.
Added to ITMS 8.5