CCX 8.x HAoWAN

[Moh.Fahmy_2]

Hi

I'm working on a design for a customer who has the below requirements:

1- 2 call center which are geographically separated (same city)

2- the 2 call centera in each site are connected through direct PSTN links

3- the 2 call centers MUST be redundant for each other (in case one fails the other should take place)

4- if one of the sites is fully loaded with calls (all agents are busy), calls should be re-routed to agents on the other site

5- if a caller that is geographically associated with the 1st call center call the second one, the call must be re-routed to the first call center.

based on this requirements, i thought of proposing UCCX 8.x with HA over WAN. the two agents groups will be differentiated using skill groups. and from the perspective of the active UCCX server, a group will be local and the other will be remote. if WAN is down, the remote agents will fail over to the local standby UCCX server. and if the active UCCX server is down, the two agent groups will failover to the standby CCX server.

So point 3 will be satisfied.

However, i needed to know if points 4 and 5 could be staified using this setup or not. and if yes HOW?

thanks in advance.

Ahmed Esmat.

[Jonathan Schulenberg · Accepted solution]

4- if one of the sites is fully loaded with calls (all agents are busy), calls should be re-routed to agents on the other site

The answer here depends on what particular part of the system would be busy:

• If you have run out of agents at site A and there are agents at site B assigned to the skill (at a lower competency rating) and available, CCX will automatically offer the call to them. Do not use resource groups; always use skills-based routing.
• If you have run out of PSTN call paths at site A, the carrier would need to reroute these calls to circuits at site B. How they would do this will vary by carrier and whether or not both sites are in the same rate center. One way or another, the carrier needs to do this part. As long as there is sufficient WAN bandwidth to connect the call back to the CCX node at site A, the call should proceed as normal.
• If you have maxed out the CTI Ports on the active CCX server at site A, you're done and you should be quoting UCCE instead.

None of these account for equipment failures on your part. You'll still want to think through PSTN gateway or UCM node failures. Also remember that there is no way to state that the CCX node at site A is preferred to be active. The customer will need to manually fail back to the site A node.

5- if a caller that is geographically associated with the 1st call center call the second one, the call must be re-routed to the first call center.

Your answer doesn't identify how the caller is associated with one site or another. Also, this seems like a useless objective because they will be handled by the same infrastructure at either site. Here are three guesses though:

• If using the callers area code (doesn't account for international calls or unknown caller ID calls), the script will need to query a database or XML file to determine what CSQ to place the call in.
• If using the called number, the call cannot arrive at the "wrong site" unless you are having a failure as described above and the carrier sent it there.
• If using an account number or another caller-provided identifier, the script will need to query a database or XML file to determine what CSQ to place the call in.

If for some reason you must move the caller from a circuit at site X to a circuit at site Y, you could use a Call Redirect step within the script and provide a DID for the other site. This seems pointless, will add delay/ringback to the callers experience, and will artificially inflate the reports in CCX.

[Ahmed Khalefa]

Perfect answer and analysis .

[Moh.Fahmy_2]

Thanks a million Jonathan that's very helpful.