Superclustering
Polycom, Inc. 228
● Add the system to an existing supercluster by pointing it to one of the existing clusters in the
supercluster. The system you’re logged into becomes one of the clusters in that supercluster, and its
local data store is largely replaced by a copy of the shared supercluster data store.
If a supercluster exists, the Remove from Supercluster command lets you remove the cluster selected in
the list from the supercluster, re-initializing it as a new stand-alone cluster. It retains the data and
configuration from the supercluster (including site topology), but that data is no longer synchronized to the
common data store. If the cluster you’re removing is responsible for any territories (as primary or backup),
you must first reassign those territories. The cluster being removed may be either the one you’re logged into
or another cluster. The system prompts you to confirm.
The Busy Out command gracefully winds down the use of the selected cluster:
● Existing calls and conferences on the selected cluster continue, but no new conferences are allowed
to start. New calls are allowed to start only if they are associated with existing conferences.
Registrations are rejected, except for endpoints currently involved in calls. The cluster ceases to
manage bandwidth.
● Territories for which the selected cluster has primary responsibility and a different cluster has backup
responsibility are transferred to the backup cluster.
● Registrations are seamlessly transferred to the backup cluster (for endpoints that support this).
Bandwidth usage data for ongoing calls is seamlessly transferred to the backup cluster.
The Stop Using command takes the selected cluster immediately out of service:
● Existing calls and conferences on the selected cluster are disconnected. No new calls or conferences
are allowed to start. All registrations are rejected. The cluster ceases to manage bandwidth.
● Territories for which the selected cluster has primary responsibility and a different cluster has backup
responsibility are transferred to the backup cluster.
● Registrations are seamlessly transferred to the backup cluster (for endpoints that support this).
Bandwidth usage data for ongoing calls is seamlessly transferred to the backup cluster.
Caution: Adding a Cluster to a Supercluster Overwrites Data
When you add the cluster you’re logged into to an existing supercluster, virtually all of that cluster’s
data and configuration are replaced by the shared data and configuration of the supercluster. This
includes, among other things, users, groups, conference rooms, site topology, Conference Manager
configuration, Call Server configuration, and integrations.
When you create a new supercluster, the data and configuration of the cluster you’re logged into are
replaced by the data and configuration of the cluster to which you’re pointing it.
Be sure you create a new supercluster by joining the cluster you’re logged into to the cluster that has
the data and configuration you want to preserve. For instance, if one of the clusters is integrated with
your Polycom RealPresence Resource Manager or CMA system, join the other cluster to it, not the
other way around.
Note: Superclusters and Resource Management Integration
You can’t add a Polycom RealPresence Resource Manager or CMA system to a supercluster or
create a supercluster with a Polycom RealPresence Resource Manager or CMA system. But you can
integrate a Polycom RealPresence DMA cluster with a Polycom RealPresence Resource Manager or
CMA system in order to get site topology and user-to-device association data from the latter (see
Resource Management System Integration on page 178). You can do this either before or after
creating a Polycom RealPresence DMA supercluster. The site topology and user-to-device
association data is replicated throughout the supercluster.