CCM 5.1 MGCP Gateway Registration Problem

[keithknowles]

I have 3 CCM servers, 1 publisher, and 2 subscribers. I also have 3 MGCP PRI gateways. Ideally I would like to register the gateways between my two subscribers so that if one sub failed not all the gateways would. The problem I am seeing is that my gateways will register to the publisher, and the primary subscriber just fine, but they will not register to my secondary subscriber. Initially this worked just fine, then we renamed, added, removed, and reconfigured our partitions and CSS. Once that was done we rebooted the CallManagers and then this happened. I have even tried adding a new gateway to the secondary subscriber without success. If i do a "show mgcp" on the gateways when they are pointed to the secondary subscriber, they say they are registered even though CCM says they are not, this I think is simply the keepalive/signaling port though and not a true registration. If anyone has any info on what I am seeing here I would really appreciate it, thanks alot in advance.

Keith

[Adrian Saavedra]

Hi,

How is your ccm-manager redundant-host command configure in your gateways?

Regards,

-- adrián.

[keithknowles]

ccm-manager redundant-host

[Adrian Saavedra]

Hi,

Also you can do a show ccm-manager in your gateways to see the Cisco CallManager failover options.

Regards,

-- adrián.

[keithknowles]

bruce#sh ccm-mana

MGCP Domain Name:

Priority Status Host

============================================================

Primary Registered

First Backup Backup Ready

Second Backup Backup Ready

Current active Call Manager:

Backhaul/Redundant link port: 2428

Failover Interval: 30 seconds

Keepalive Interval: 15 seconds

Last keepalive sent: 11:19:54 MDT Mar 19 2008 (elapsed time: 00:04:34)

Last MGCP traffic time: 11:24:27 MDT Mar 19 2008 (elapsed time: 00:00:01)

Last failover time: 08:09:26 MDT Mar 17 2008 from (142.66.4.21)

Last switchback time: 08:09:43 MDT Mar 17 2008 from (142.66.4.22)

Switchback mode: Graceful

MGCP Fallback mode: Not Selected

Last MGCP Fallback start time: None

Last MGCP Fallback end time: None

MGCP Download Tones: Disabled

TFTP retry count to shut Ports: 2

Backhaul Link info:

Link Protocol: TCP

Remote Port Number: 2428

Remote IP Address:

Current Link State: OPEN

Statistics:

Packets recvd: 78649

Recv failures: 0

Packets xmitted: 65844

Xmit failures: 0

PRI Ports being backhauled:

Slot 0, VIC 0, port 1

Slot 0, VIC 0, port 0

Configuration Auto-Download Information

=======================================

Current version-id: 1201035740-e1d6628f-fbff-482b-b245-5678bbe51a93

Last config-downloaded:00:00:00

Current state: Waiting for commands

Configuration Download statistics:

Download Attempted : 1

Download Successful : 1

Download Failed : 0

TFTP Download Failed : 0

Configuration Attempted : 1

Configuration Successful : 1

Configuration Failed(Parsing): 0

Configuration Failed(config) : 0

Last config download command: New Registration

Configuration Error History:

FAX mode: cisco

[Adrian Saavedra · Accepted solution]

Hi,

Your MGCP gateway shows that is registered to Cisco CallManager, but when you use an MGCP gateway in Cisco CallManager, sometimes it shows as unregistered in the Cisco CallManager Administration page.

Sometimes this can be because the Real-time Information Server (RIS) data collector service does not work as expected. Restart the RIS data collector service on the Cisco CallManager Publisher and Subscriber to resolve this issue.

Complete these steps to restart the RIS data collector service:

1. Go to the Cisco CallManager Serviceability page from Application > Cisco CallManager Serviceability.

2. Select Tools > Control Center.

3. Select the IP address of the server that should be restarted, then choose Cisco RIS Data collector and click Restart.

Regards,

-- adrián.

[keithknowles]

I will give that a try in the AM and let you know if it helped. Thanks Adrian.

[keithknowles]

I probably should ask though, what effect is this going to have on the CallManager/Unity/IPCC environment? Is there any particular order in which I should restart the services pub then sub or sub then pub? This is a production system with some 500 people on it and 1500 running through the gateways so I have to be very careful when making system changes.

Thanks again Adrian.

[Jaime Valencia]

RIS restart won't affect anything of that

verify the domain name under CUCM to be:

hostname + domain(if configured)

How to Configure MGCP with Digital PRI and Cisco CallManager

http://www.cisco.com/en/US/partner/tech/tk1077/technologies_configuration_example09186a00801ad22f.shtml

Configuring the Cisco IOS MGCP Gateway

http://www.cisco.com/en/US/partner/products/sw/voicesw/ps556/products_tech_note09186a008017787b.shtml

HTH

javalenc

if this helps, please rate

[keithknowles]

We are using IP addresses in place of host names to remove DNS dependancy.

I have verified the configuration of the router, and it will register with any other CCM server I point it to, just not the secondary production subscriber.

I think that Adrian is onto something with the RIS restart though. The gateways "show mgcp" and "show ccm" outputs display the gateway as registered even though CUCM Admin does not.

[Jaime Valencia]

that message only means that keepalives at port 2428 are being sent and received, not necessarily that the HW is indeed registered and working

[keithknowles]

Thats true. But if a change the mgcp call agent to my publisher it registers, change it to my primary sub it registers, change it to my test publisher it registers, test sub also registers, just not to my secondary sub. This tells me that the router is properly configured (or it wouldn't register to anything) and that the issue lies with my secondary subscriber.

[keithknowles]

Adrian you rock!!! That worked perfectly. Thanks alot.