Default configuration policy automatically always updates to all computers. You should never have to deploy it anywhere. As long as the Enterprise agent is correctly pointing to one of the DM's or one of the DM's Scalability Servers, it should receive that policy automatically as well.
Depending on the system, possible issues could be:
1. Name resolution in the reverse direction (run 'camping -a <other machine name>' from each end of a connection to test)
2. UDP instability with larger chunks of data (run 'camping -s 20000 <other machine name> from each end to test)
3. Something hung on the SS or Client side (reboot each to confirm)
4. A damaged installation (Record any patches applied, run a repair, reboot, apply patches and reboot once more)
Barring that, CCSMAPI and CCSMACT on the DM side can be recycled if the DM cannot be rebooted, and any logs with those terms in the name as well as CCNFAGENT, CCNFAGENTAPI and CCSMAGENT should be checked/collected where they exist on the DM/SS/Client side (they don't all exist on all levels) to look for a deeper cause.
Thanks,
Joe Puccio
Senior Engineer, Software Support