https://fedorahosted.org/389/ticket/48636

Replication latency, especially over a WAN, can become worse when there are several masters receiving updates at the same time.  What happens is that one master will take exclusive access of a replica, and not release it for a very long time.  This blocks the other masters from sending their updates to that consumer, and this adds to the replication latency as those updates have to travel back and forth with all the other masters, and consumers.  See the bugzilla for more detailed info.

We need a way to notify a master that it is holding its exclusive access of a replica for too long, and that it needs to yield so other masters can start sending some of their updates to that replica.

Design doc: http://www.port389.org/docs/389ds/design/repl-conv-design.html
Implementation
The new configuration attribute can be set using ldapmodify, and the change takes effect immediately. It does not require a server restart.

Major configuration options and enablement
The new configuration attribute is: nsds5ReplicaReleaseTimeout, and it accepts any value of zero or higher. This value represents a timeout in seconds. This attribute is set in each replica mapping tree entry. It should be set on Masters and Hubs - there is no need to set this on consumers. Setting the timeout to zero is the same as disabling the timeout feature. Below is an example of setting a timeout.

dn: cn=replica,cn=dc\3Dexample\2Cdc\3Dcom,cn=mapping tree,cn=config
changetype: modify
replace:  nsds5ReplicaReleaseTimeout
nsds5ReplicaReleaseTimeout: 60
Internal testing has shown that 60 seconds seems to be the ideal value. Any value below 30 seconds has an adverse impact on replication performance.