PCCE - Queued Calls Disconnect after 1.5 Hours After MaxTimeinQueue Raised to 4 Hours

[kyle_hanson]

Our call center is slammed and calls are holding longer than normal. We ran into the usual hour limit of the MaxTimeinQueue and raised it to 2 and then 4 hours. However, calls still seem to be disconnected regularly around 1.5 hours, well below our setting. They receive a message that they were no longer in the queue and should hang up and call again. I also seem to have trouble locating the calls in my usual reports and logs. Could something else be going on? Is there a hard limit to queue time, or are they reaching some other limit?

PCCE version 11.6.1

[Omar Deen]

Can you dump and attach the rtr log from the Call Router? Do you have a lot of nodes that your callers go through - especially during queue treatment?

[kyle_hanson]

I was curious if they may be hitting a "too many nodes" limit, but can't say for sure. We don't use a ton of nodes while holding, but calls didn't used to hang this long. They get 30 seconds of music per node, so that would add up if that's what you're thinking. I have attached the rtr logs for A and B for the past day. I can't give a time when one of these happened due to lack of info from my users. If you can find anything though I'd appreciate it. 

[Omar Deen · Accepted solution]

You're hitting your max nodes limit

10:31:28:496 ra-rtr Script: Default\\New_TS executed for more that: 1000 nodes. 

That was in the rtr log about 12 times. So you'll need to increase your max node count. Log into your Router/Rogger, open the registry and navigate to HKEY_LOCAL_MACHINE\SOFTWARE\Cisco Systems, Inc.\ICM\<your_instance>\RouterA\Router\CurrentVersion\Configuration\Script

From there, you'll see the MaxNodesExecuted key. This number should be an educated guess based on your queue times. If you have long queue times, and it looks like you do, I'd jack this up to like 2000

[bill.king1]

Thank you for posting this.

I'm guessing this is a dynamic setting and once you update it, no cycling of services are needed?

And is there any documented impact/caveat from Cisco of increasing this setting as far as capacity reduction (aside from the risk of potentially maxing out your VXML licenses).

[Omar Deen]

Hi Bill... there's always going to be an inherent risk when you increase the MaxNode limit. The way the Router/Rogger is sized these days, there should be no issues at all with increasing that to 1500. Once you go above that, you'll want to monitor memory utilization. If memory utilization goes above 95%, you'll want to consider adding more memory to the VM guest (it is supported). I've increased that to 2000 in several environments with no problems - but of course, your mileage may vary.

As for documentation, I'm not aware of any Field Notices, Troubleshooting Tech Notes or Security Advisories. Most of the time, this reg key will never be touched, regardless of how large your call center is because this is all about routing and being efficient with your nodes. Example... using nested IF statements rather than a bunch of a IF nodes.