cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1997
Views
0
Helpful
5
Replies

UCCX HA Deployment configuration best practise

d.vieritz
Level 1
Level 1

I am about to deploy UCCX 11 HA over WAN into an existing CUCM 11 environment.  Customer has 2 Data Centres for the servers.  The customer has 2 x CUCM and 2 x Unity Connection Servers.   The CUCM and Unity Connection Publisher Servers are deployed in the DR Data Centre so that it is possible to make Config changes in a DR situation when the primary DC is down.  The CUCM and Unity Connection subscriber/main call processing servers are located in the Primary DC which is closer to all the users for optimum performance. Is it possible to deploy UCCX so that the First Node/Publisher is in the DR Data Centre and the Second Node is in the Primary DC and is running as the Master and actively handling the Contact Centre processing?.   This would mean that during normal operation all configuration work will be performed on the CUCM/UCXN/UCCX Servers in the DR Data Centre and all call processing would be occurring in the Primary DC.    

I know that UCCX Config changes are not allowed in a DR situation when one UCCX HA node is down, but at least during normal operation we would know that all configuration work occurs on the Servers in the DR Data Centre and all call processing would be occurring in the primary DC.   By default the UCCX First Node/Publisher is the Master.  Is there any issue with making the second node run as the Master?.

5 Replies 5

Deepak Rawat
Cisco Employee
Cisco Employee

There is no harm in making the secondary UCCX Node run as Master, lot of customers do that basis on their requirement (specially if the users/agents are close to that location) and it is absolutely supported. However your below statement is not valid for UCCX since it works on the concept of Active-Standby opposed to CUCM/UCxN that works on the concept of Active-Active

 "during normal operation we would know that all configuration work occurs on the Servers in the DR Data Centre and all call processing would be occurring in the primary DC."

What I mean by Active-Standby is that at a given point of time everything will work through one server only that is Master. Be it call routing, agent login, prompts playing etc whatever you can think of from UCCX perspective. As far as configuration is concerned that can be done from any server in a HA environment.

Regards

Deepak

Thanks Deepak,

Sounds good. I understand that UCCX is an Active/Standby system so if the Node 2 in the Primary DC is the Master then all IVR/Contact Centre processing will occur on Node 2.  As I understand it Node 1 (The DB Publisher and Slave server) would be used to run Historical reporting.  Is that correct?.  We could also point our WEB browser at either server to perform configuration activities so that is good.

I understand that UCCX is an Active/Standby system so if the Node 2 in the Primary DC is the Master then all IVR/Contact Centre processing will occur on Node 2.

Answer :Your understanding is absolutely correct

As I understand it Node 1 (The DB Publisher and Slave server) would be used to run Historical reporting.  Is that correct?.

Answer : Correct, reports will always be pulled from the Non Master node. So if Node 1 is acting as Slave reports will be pulled through that. If Node 2 that will be installed in DR site is acting as Slave, then the HR reports will be pulled against that. If only one server is available in the HA node then of course reports will be pulled against that only.

We could also point our WEB browser at either server to perform configuration activities so that is good.

Answer : There are two aspects to this:

1) Real Time Reports/Activities that customers want to show on the big screens that are deployed in Contact Centers commonly known as Wallboards. In that case, you will need to point the Wallboard connection to the server which is running as Master since the Real Time stats are only published/updated on the master node

2) For custom reporting or any other thing that is not being done in real time, you should always point it towards the slave node so that the load on Master Node is as minimal as possible.

Regards

Deepak

Hi Deepak,

Thanks.......one last question re the WEB browser.  When we run Appadmin to configure CSQs Resources etc can we point the Appadmin Browser at either node.  Our plan would be to point it at the Node 1 (running as slave) in the DR Data Centre.  Any configuration change we make would be to add resources, queues etc which would immediately apply to the processing being performed by Node 2 (running as Master) in the Primary Data Centre.  Is this a recommended/good approach?.

Cheers,

David

Ok from web browser you mean CCX Administration, sorry I assumed it from a third party perspective. Absolutely, you can use any server to configure / add / remove changes within the CCX Environment. Even though Node 1 will be running as Slave you can still use it to make the changes and ideally you should only use that as a preference since even though it is acting as Slave but the DB entry for that is still 1 and will remain 1. Changes will replicate to the second node running as Master instantly. This will work vice versa as well.

Regards

Deepak