redundancy group is comprised of two
IM and Presence Service nodes from the same cluster. Each
node in the presence redundancy group monitors the status, or heartbeat, of the peer node. You can configure a presence redundancy group to provide both redundancy and recovery for
IM and Presence Service clients and applications.
Failover—Occurs in a presence redundancy group when one or more critical services fails
IM and Presence
Service node in the group or a node in the group fails. Clients
automatically connect to the other
IM and Presence
Service node in that group.
Fallback—Occurs when a fallback command is issued from the CLI or Cisco Unified Communications Manager during either of these conditions:
IM and Presence
Service node comes back into service and all critical services are
running. The failed-over clients in that group reconnect with the recovered
node when it becomes available.
IM and Presence
Service node fails due to a critical service failure, and the peer
node is in the Failed Over state and supports the automatic recovery fallback.
For example, if you are using
presence redundancy groups, Cisco Jabber clients will fail over to a backup
IM and Presence Service node if the services or
hardware fail on the local
IM and Presence Service node. When the failed node
comes online again, the clients automatically reconnect to the local
IM and Presence Service node if you have configured automatic fallback. If you have not configured automatic fallback, you can manually initiate the fallback when the failed node
In addition to redundancy and recovery, presence redundancy groups also allow you to configure high availability for your cluster.
The IM and Presence Service supports high availability for multiple-node deployments.
After you configure a
presence redundancy group, you can enable high availability for
the group. A pair
of nodes is required for high availability. Each node has an
independent database and set of users operating with a shared
availability database that is able to support common users.
All IM and Presence
Service nodes must belong to a presence redundancy group,
which can consist of a single IM and Presence Service
node or a pair of IM and Presence Service nodes.
You can configure high availability using two different modes:
Balanced mode: This mode provides redundant high availability with automatic user load balancing and
user failover in the event that one nodes fails because of component
failure or power outage.
Active/standby mode: The standby
node automatically takes over for the active node if the active
node fails. It does not provide automatic load balancing.
We recommend that you configure your IM and Presence Service deployments as high availability deployments. Although you are permitted to have both high availability and non-high availability presence redundancy groups configured in a single deployment, this configuration is not recommended.
Presence Redundancy Group Prerequisites
For deployments over the WAN, a minimum of 10 megabits per second of dedicated bandwidth is required for each IM and Presence Service cluster, with no more than an 80-millisecond round-trip latency. Any bandwidth less than this recommendation can adversely impact performance.
Redundancy Group Task Flow
IM and Presence Service node can be assigned to only
one presence redundancy group. For high availability, you must assign two nodes
from the same cluster to the presence redundancy group and enable high
availability for the group.
critical services are running on the nodes that you plan to add to a presence redundancy group. Critical services must be running before you
turn on high availability. If critical services are not running
on either node, the presence redundancy group will go into a Failed state when you turn on
high availability. If critical services are not running on one node, then that
node fails over to the other node when you turn on high availability.
Use Cisco Unified Communications Manager to configure redundancy for IM and Presence Service nodes.
Each presence redundancy group can contain two IM and Presence Service nodes. Each node can be assigned to only one presence redundancy group. Both nodes in the presence redundancy group must be on the same cluster and have the same IM and Presence Service database publisher node.
Configure optional service parameters that determine the keep alive settings by which each peer in a presence redundancy group monitors the heartbeat (i.e., the status) of its peer node in order to confirm that the peer is active. A failover can be initiated if the peer node is unresponsive after a configured timer expires.
Cisco recommends that you use the default values for these service parameters. However, you can also reconfigure the values to suit your needs.
In Cisco Unified CM IM and Presence Administration, choose System > Service Parameters.
From the Server drop-down, select an IM and Presence node
From the Service drop-down, select Cisco Server Recovery Manager (Active).
Under General Server Recovery Manager Parameters (Clusterwide), configure the clusterwide Keep Alive settings that each node in a Presence Redundancy Group uses to monitor monitor the heartbeat of its peer node. A failover can be initiated if the peer node is unresponsive.
Service Port— This parameter specifies the port that Cisco Server Recovery Manager uses to communicate with its peer. The default is 22001.
Admin RPC Port—This parameter specifies the port that Cisco Server Recovery Manager uses to provide admin rpc requests. The default is 20075.
Critical Service Delay—This parameter specifies the duration in seconds that a critical service can be down before failover is initiated. The default is 90.
Enable Automatic Fallback—This parameter specifies whether to do automatic fallback. In the event of a failover, the IM and Presence Service moves users automatically from the backup node to the primary node thirty minutes after the primary node returns to a healthy state. The default value is False.
Initialization Keep Alive (Heartbeat) Timeout—This parameter specifies the duration in seconds that the heartbeat can be lost with the peer during initialization before failover is initiated. The default is 120.
Keep Alive (Heartbeat) Timeout—This parameter specifies the duration in seconds that the heartbeat can be lost with the peer before failover is initiated. the default is 60.
Keep Alive (HeartBeat) Interval—This parameter specifies the interval in seconds between keep alive (heart beat) messages being sent to the peer. The default is 15.
Configure the following additional parameters, which tell CUPC 8.5 and higher clients how long to wait before attempting to relogin. Unlike the above parameters, these parameters must be configured separately for each cluster node.
Client Re-Login Lower Limit—This parameter specifies the minimum number of seconds which CUPC 8.5 (and higher) should wait before attempting to re-login to this server. The default is 120.
Client Re-Login Upper Limit—This parameter specifies the maximum number of seconds which CUPC 8.5 (and higher) should wait before attempting to re-login to this server. The default is 537.
What to Do Next
If you did not enable high availability when you configured the presence redundancy group, Enable High Availability now.
Enable High Availability
Failure to set up replication in the
IM and Presence Service cluster and ensure that all critical services are running may result in an immediate failover when high availability is enabled for the presence redundancy group.
Ensure that replication is set up in the
IM and Presence Service cluster.
Ensure that all critical services are running.
From Cisco Unified CM Administration, choose System > Presence Redundancy Groups.
Specify search criteria and then click Find.
Choose the presence redundancy group that you configured.
To enable high availability, check the Enable High Availability check box.
Configure User Assignment Mode
Use this procedure to configure the way in which the sync agent distributes users to the nodes in the cluster. This setting helps to manage failover and load balancing.
From Cisco Unified CM Administration, choose System > Enterprise Parameters.
In the User Management Parameters Area, choose one of the following options for the User Assignment Mode for Presence Server parameter:
Balanced—This mode assigns
users equally to each node in each subcluster and attempts to
balance the total number of users equally across each node. This is the default option.
Active-Standby—This mode assigns all users to the first node of the
subcluster, leaving the secondary server as a backup.
None—This mode results in no assignment of the users to the nodes in the cluster
by the sync agent.
Redundancy Interactions and Restrictions
Multiple Device Messaging
The Multiple Device Messaging feature causes a delay with server recovery on the IM and Presence Service if failover occurs. If server failover occurs on a system where Multiple Device Messaging is configured, the failover times generally are twice as long as the times specified with the Cisco Server Recovery Manager service parameters.