Can UCCX be integarted with multiple CUCM clusters

Unanswered Question
May 26th, 2008

Hi All,

Can a UCCX deployment be integarted with multiple CUCM clusters or only one CUCM cluster


I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
lpezj Mon, 05/26/2008 - 00:57


I think that UCCX can only be integrated with one Callmanager cluster. It is likely that you can set up agents is one cluster and send calls to/from one cluster to another, but all the agents must be in the same cluster.

You can use two UCCX, one for each cluster, and use IPCC Enterprise to load balance between the UCCX agents in each cluster.

You can too use only IPCC Enterprise with one PG for each cluster to load balance or distribute calls among the two CallManager clusters.

Hope this helps,

Juan Luis

rbua Mon, 05/26/2008 - 04:54


you will need to define an application and profile on each cluster, there is no mean to load balance across clusters or applications, so agents will need to be used by one or the other, licenses will be the ones in use, you might eventually spare some hardware, but the overall administration will be quite cumbersome to say the least(make scripts twice and so on...).

Using UCCE you might be able to balance the load across ipccx and clusters as Juan Luis was rightfully suggesting.



Chris Deren Mon, 05/26/2008 - 08:13

ONLY ONE cluster, as you can only point one CM at a time via AXL on UCCX 4.5/5.0 or directory on 4.0 and earier. So, if you need to point to another cluster you will need to recreate the integraion.

You can point multiple UCCX environments to a single CM cluster.


agtmcgarry Thu, 06/26/2008 - 01:46


I have




I have two callmanager clusters.

I created a generic pg and configured both callmanager clusters and cvp as pims 1,2,3

All pims went active and all looked good.

However I found that I could only logon to the first callmanager. On the jgw process I could see the logon request to a device on the second callmanager cluster going to the first cluster jgw process and CTIOS logon failing.

Do you know a way of specifying the second callmanager cluster.

I also tried creating a second pg. So I deleted the second callmanager pim from the first pg, updated the routers for 2 pgs and installed a second generic pg with one callmanager pim.

Now both my jgw processes keep crashing and neither callmanager goes active.

Any help would be much appreciated.



rbua Thu, 06/26/2008 - 01:51

Hi Anthony are you using the same application, JTAPI user and call searching space or different ones?

agtmcgarry Thu, 06/26/2008 - 02:02


I have two seperate UCM clusters, both 6.1. I have two PG's with ICM7.2.4 and CTIOS7.2.4. Each cluster has an ipcc pg user configured.

They both use the same username.


rbua Thu, 06/26/2008 - 02:03

Hi Anthony,

please try using a different username and let me know if the issue persist.



Chad Stachowicz Sun, 06/29/2008 - 22:09

In my travels and my opinion. Each CTIOS instance must be the SAME PID as its associated agent PG, and if its not login's will not work. Therefore useing this same logic, each agent PG to each cluster need its own instance of CTIOS. And you then need to login to the correct CTIOS instance. This is the reason you are always seeing it try the first cluster! Just my 2 Cents.


agtmcgarry Mon, 06/30/2008 - 06:42

Hi Chad,

Yes in the CTIOS setup you have to select the PID of the ACD. Therefore in my setup with a side A and B CTIOS both where using the PID from UCM cluster 1. Therefore all requests where going only to the first cluster.

A workaround for now was to break the side A/B and have CTIOS and PG's in simplex. One for each cluster.




This Discussion