Making calls from backup/redundant CCM?

Unanswered Question
Oct 17th, 2007

Hello!

I have 2 CCM servers configured in a server group (CCM1, CCM2) that most of my phones are registered with. I also have 2 Cisco 2851 gateways running MGCP with PRIs that the CCMs can route to. When registered to CCM1, I can make calls just fine, but when registered to CCM2, calls fail.

If I do a 'show ccm-manager hosts' from either gateway, I can see CCM1 is registered with CCM2 as a First Backup, status of "Backup Ready". Do the calls from CCM2 fail because it is only a backup and not actually registered with the gateway?

So, can both CCMs be registered to the gateways at the same time? I am concerned that if this is normal behavior, what happens if I lose IP connectivity to CCM1 from my phone, but the gateway still sees it as being alive, therefore never failing over to CCM2 so that I can make calls?

Any good documents on this are welcome. I just haven't been able to find this particular information for whatever reason... Thanks!

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
rob.huffman Wed, 10/17/2007 - 06:42

Hi Team,

Here is a doc that outlines the required steps;

Cisco IOS gateways can maintain links to up to two backup Cisco Unified CallManager servers in addition to a primary Cisco Unified CallManager. This redundancy enables a voice gateway to switchover to a backup if the gateway loses communication with the primary. The backup server takes control of the devices that are registered with the primary Cisco Unified CallManager. The second backup takes control of the registered devices if both the primary and first backup Cisco Unified CallManager fail. The gateway preserves existing connections during a switchover to a backup Cisco Unified CallManager.

When the primary Cisco Unified CallManager server becomes available again, control reverts to that server. Reverting to the primary server can occur immediately, after a configurable amount of time, or only when all connected sessions are released.

Switchback

Switchback is the process a voice gateway uses to reestablish communication with the primary Cisco Unified CallManager server when the server becomes available again. Switchback can occur immediately, at a specified time after the last active call ends, or after a specified length of time.

MGCP Gateway Fallback

The MGCP gateway maintains a remote connection to a centralized Cisco Unified CallManager cluster by sending MGCP keepalive messages to the Cisco Unified CallManager server at 15-second intervals. If the active Cisco Unified CallManager server fails to acknowledge receipt of the keepalive message within 30 seconds, the gateway attempts to switch over to the next available Cisco Unified CallManager server.

SUMMARY STEPS

1. enable

2. configure terminal

3. ccm-manager redundant-host {ip-address | DNS-name} {ip-address | DNS-name}

4. ccm-manager switchback {graceful | immediate | schedule-time hh:mm | uptime-delay minutes}

5. ccm-manager fallback-mgcp

6. call application alternate

7. exit

8. ccm-manager switchover-to-backup

From this good doc;

http://www.cisco.com/en/US/products/ps6441/products_configuration_guide_chapter09186a00805583bd.html#wp1056830

Hope this helps!

Rob

isteam Wed, 10/17/2007 - 06:51

So an MGCP gateway can only be controlled by one Callmanager (MGCP agent) at a time, regardless of the backup/redundancy settings?

AJAZ NAWAZ Wed, 10/17/2007 - 06:48

Hey isteam,

Just consider your two Callmanagers as one. Remember they are a cluster and recall that when you add a gateway on CCM, it is done once only. The configuration is not performed separately for each Callmanger server.

Important to understand that in fact all changes are done on the publisher which replicated the information to subscriber servers. Even if you are logged into the subscriber and apply a change - this is actually going to take affect on the publisher.

NOTE - IP phones communicate with CCM in order to route the call. Once the call is established both IP phones will communicate with each other directly using RTP stream. So if at this stage you lose connectivity to CCM from the phone, it will not matter. Any subsequent call will be handled by subscriber if publisher is unavailable.

does this answer your question?, if it does not please do not hesitate to expand your question further.

Ajaz

rstevenson1 Fri, 11/23/2007 - 08:49

So theoritically, if your Publisher went down, all should function normally via the subscriber?

Seem to have an issue at the moment, where if the Publisher is down, calls inbound from external sources works Ok. But we're unable to make external calls from the internal phones.

isteam Fri, 11/23/2007 - 09:03

My issue was that my MGCP gateway can only be registered to one device at a time. I was trying to register my phone to the subscriber and make calls like that, but the gateway would not accept them because it was registered to the publisher.

Not sure if this is your problem, but you might want to check your gateway to ensure your subscriber is set as a failover in case communication is lost with the publisher.

solomonthuo Thu, 02/04/2010 - 23:36

Hi

did you manage to solve your problem? i have the same issue. I have two UCM 7.1 and an MGCP gateway. when i take down the publisher, the gateway registers correctly on the gateway, however i can't make extenal calls but internal calls flows.

regards

solo

rajesh.kumar Fri, 11/23/2007 - 09:41

It seems like problem with MGCP router configuration

Pls. verify these commands

-----

ccm-manager redundant-host x.x.x.x

mgcp call-agent x.x.x.x

try

#no mgcp

#mgcp

also try to reset MGCP gateway at CCM

Actions

This Discussion