Dial delays on one call search space

[sheletal82]

Looking for guidance on troubleshooting an issue.

We have three Calling Search Spaces (A, B, and C) configured in our CUCM 11.5.1 environment and one of them (A) intermittently produces a delay of around 7 seconds when dialing both internal and external numbers.

The only difference I can see between the 3 of these CSS's is that "A" has about 1/3 of devices assigned to it as do the other CSS's and It has the directory URI partition selected whereas the other CSS's do not.  Could that extra partition somehow cause the delay?

What else should I investigate?

[Jaime Valencia]

Delay is most usually caused by dial plan overlapping, you need to find out exactly what is dialed when that happens, and then review your dial plan for overlapping.

You can use DNA for this.

[sheletal82]

I'm not sure what I should be looking for here but it appears to route.  Thoughts?

 

[sheletal82]

Odd thing I noticed, this only seems to effect our 8845 models.

I just programmed a 8945 with the same extension and CSS and calls go through immediately.

[Andrew West]

The call in the DNA trace is hitting a Jabber device. 

Is that applicable for all calls to Jabber devices and if so how are they connected? 

Also looks like you already took the next step I was going to say which is eliminate the Jabber side and put the DN directly on a phone. 

[sheletal82]

Yeah, 8503 is assigned to 3 devices, and 8845 deskphone and jabber on windows and android.

I just removed 8503 from the jabber devices and still have the same issue.  I also moved 8503 to a different model of phone - a 8945 - and still have the same issue.

So, the only constant here is that 8845's are the only phone that have a dial delay and only when assigned to one of our three call search spaces.

Another note worth mentioning - if I call back the other direction TO the model 8845, there is no delay.

[sheletal82]

Looks like this will become a case for TAC

[sheletal82]

This turned out to be caused by a route pattern of !*!*!*! for ip routing.  This pattern was quarantined and the problem resolved.  The indicator was the interdigit timeout triggering as indicated in the DNA