Resolution:
"_RVCM.>" messages are exchanged at start-up, termination, and loss recovery situations.  They are used by RVCM endpoints when they start or restart to find each other.

A client will publish these "_RVCM.SEARCH" messages to 'find' its clients on a per-transport basis.  Without finding its 'partner' on the other side of the CM contract, you cannot assure delivery.  This is why they are required.

Restarts of publishers and restarts of RVRDs between the clients are examples of the kinds of events that can cause these messages to be delivered.  If a source IP or a target IP address changes, if the route between hosts across RVRDs change, or publisher/subscriber restart you must expect these messages.

To avoid the flooding of "_RVCM.SEARCH" messages problem, some suggestions are listed below:

1. Make sure you are NOT running more than one CM Subscriber with the same CM Name (good to check from the running process list in your system).

2. Your CM listener applications should always be restarted with the same UNIQUE CM Name as the previous session.

3. When your CMlistener application is shutdown for a long period of time, you should remove the pre-registered listener explicitly by calling TibrvCmTransport::removeListener.

4. Remove obsolete pre-registered listeners from the ledger.

5. Only import/export "_RVCM.>" within RVRDs when necessary.

What are the possible reasons for _RVCM.SEARCH. messages flooding in the network?