Cisco Cisco TelePresence Management Suite (TMS) Version 15 Maintenance Manual
n
If the PrimaryServerId field for a given task is empty, the task will be executed by the first node that polls
the database for tasks after the task is scheduled to start.
the database for tasks after the task is scheduled to start.
n
If the PrimaryServerId field for a given task shows a node that is operational, that node will pick up the task
as soon as the task is scheduled to start.
as soon as the task is scheduled to start.
n
If the PrimaryServerId field for a given task shows a node that is not operational, the other node will pick up
the task after 60 seconds and execute it.
the task after 60 seconds and execute it.
Cisco TMS does not directly assign tasks to nodes based on server load. The other node picks up a task if
the original node is under such a heavy load that it is unable to poll the database for new tasks—this offers a
similar balancing mechanism.
the original node is under such a heavy load that it is unable to poll the database for new tasks—this offers a
similar balancing mechanism.
The other node will not pick up and re-execute the task if the original assignee fails during execution of that
task.
task.
Live Service and the Conference Control Center
When a conference is booked in Cisco TMS, it is allocated to the Live Service on one of the two nodes. The
Live Service collects data about the conference, which will display in
Live Service collects data about the conference, which will display in
Monitoring > Conference Control
Center
in Cisco TMS.
If a conference is owned by the Live Service on tms01, but the load balancer forwards feedback from one of
the participants in the conference to tms02, tms02 uses a remote procedure call (RPC) to pass the
information to tms01.
the participants in the conference to tms02, tms02 uses a remote procedure call (RPC) to pass the
information to tms01.
If a user logged on to tms01 makes a change to an ongoing conference owned by tms02 in
Conference
Control Center
(CCC):
1. Tms01 sends the change to tms02 using RPC.
2. Tms02’s Live Service then instructs the systems involved to make the change.
Live Service uses TCP port 8085 for RPC. If changes made in CCC are slow to propagate from one node to
the other:
the other:
1. Check that a firewall is not blocking traffic on port 8085.
2. Verify that both Cisco TMS servers are listening on port 8085:
l
Open a cmd shell and execute a netstat -a -n | findstr :8085 command on both servers
l
Try to connect using telnet (telnet tms01.example.com 8085) and check that the connections
are not refused .
are not refused .
To check which node owns an ongoing conference and its participants, see the RunningServer column of the
LiveStatusData table in the tmsng database. This field contains either “tms01” or “tms02”.
LiveStatusData table in the tmsng database. This field contains either “tms01” or “tms02”.
Cisco TMS Administrator Guide (14.1)
Page 288 of 320
Redundant deployments
Deploying with a load balancer