I have a single server (let’s call it “Webmail”) that holds replicas of all the users’ mail databases in my domain. I also have a cluster (“Mailcluster”) of servers that hold replicas of those mail files.
My Domino Directory has a connection document that causes Webmail to initiate push/pull replication of the mail\ folder with Mailcluster. My Domino Directory does not contain any connection document that specifies replication between Webmail and any single member of Mailcluster; only the aforementioned connection with the cluster.
As expected, Webmail initiates scheduled replication with Mailcluster. I see in Webmail’s log replication event documents that correspond with this. However, I also see replication event documents there indicating replication with one of the specific servers in the cluster. In other words, I see some replication events categorized under “Mailcluster”, and others categorized under the names of the servers in Mailcluster.
What I think is happening is that the Webmail begins replicating with one of the servers in Mailcluster (as expected). At a certain point, the info pertaining to the replication event is reaching a size limit (approx 40KB – there are 800+ mail files), so it writes the replication event out to disk, and carries on replicating. When it finally finishes, it writes out the remaining info, but this time it puts the name of the specific server in the “sourceserver” field (which is the one used for categorization in the view).
Any body else seeing similar behavior?
Thanks - Steve