Cisco Cisco TelePresence Video Communication Server Expressway 관리 매뉴얼
■
A SIP call between the endpoints consumes two traversal call licenses (one on each node).
■
An H.323 call between the endpoints consumes two traversal call licenses (one on each node).
Capacity alarms are raised if either of the following usage thresholds are reached:
■
the number of concurrent traversal/non-traversal calls reaches 90% of the capacity of the cluster
■
the number of concurrent traversal/non-traversal calls on any one unit reaches 90% of the physical capacity
of the unit
of the unit
Example deployment
If, for example, you want to deploy a resilient cluster that can handle up to 750 concurrent non-traversal calls and
250 concurrent traversal calls you could configure 4 peers as follows:
250 concurrent traversal calls you could configure 4 peers as follows:
Peer 1
Peer 2
Peer 3
Peer 4
Total cluster
capacity
Non-traversal call licenses
250
250
250
0
750
Traversal call licenses
100
100
50
0
250
It would not matter to which peer an endpoint registers as the call licenses are shared across all of the peers. If any
one of the peers is temporarily taken out of service the full set of call licenses will remain available to the entire
cluster.
one of the peers is temporarily taken out of service the full set of call licenses will remain available to the entire
cluster.
However, we recommend that, where possible, the number of licenses is configured evenly across all peers in the
cluster.
cluster.
Managing Clusters and Peers
Setting Up a Cluster
Prerequisites
Before setting up a cluster of X8.7.2 VCS peers or adding an X8.7.2 VCS to a cluster, ensure that:
Platform and Software Versions Match
■
All clusters peers are running the same version of code. The only occasion where different peers are allowed
to run different versions of code is for the short period of time while a cluster is being upgraded from one
version of code to another, during which time the cluster will operate in a partitioned fashion.
to run different versions of code is for the short period of time while a cluster is being upgraded from one
version of code to another, during which time the cluster will operate in a partitioned fashion.
■
Each peer is using a hardware platform (appliance or virtual machine) with equivalent capabilities; for
example, you can cluster peers that are running on standard appliances with peers running on 2 core Medium
VMs, but you cannot cluster a peer running on a standard appliance with peers running on 8 core Large VMs.
example, you can cluster peers that are running on standard appliances with peers running on 2 core Medium
VMs, but you cannot cluster a peer running on a standard appliance with peers running on 8 core Large VMs.
Network Conditions Are Met
■
Each peer has a different LAN configuration (a different IPv4 address and a different IPv6 address, where
enabled).
enabled).
■
Each peer in a cluster is within a 15ms hop (30ms round trip delay) of each and every other VCS in or to be
added to the cluster.
added to the cluster.
■
Each peer in a cluster is directly routable to each and every other VCS in or to be added to the cluster. (There
must be no NAT between cluster peers – if there is a firewall ensure that the required ports are opened.)
must be no NAT between cluster peers – if there is a firewall ensure that the required ports are opened.)
178
Cisco TelePresence Video Communication Server Administrator Guide
Clustering and Peers