Conceptually for MMR, every node is both a provider and consumer, so
it is configured both to be a master and a replica at the same time. ServerIDs are used
to >distinguish which changes occurred on which master, and to avoid replication loops.
That is independent of which replication mechanism you choose to use between >the
nodes (i.e., this has to be done regardless of syncrepl or delta-syncrepl).
Okay I think I'm getting it.
So I have a working config on two lab vm.
It is syncing the acesslog changes, but when I fired up the second server, it didn't
do a full sync.
I need multiple RID's for each item I want to sync?
I'm doing this all in slapd.conf