[cisco-voip] E164 routing loop

Ki Wi kiwi.voice at gmail.com
Sun Aug 13 22:47:30 EDT 2017


Hi Group,
I have encountered this interesting problem on customer PBX. Didn't work on
live system for a long time but I am pretty sure this shouldn't be a
default behavior.

When external PSTN caller calls an unassigned number in the DID range, CUCM
returns with error code 27 ( destination out of order).

This causes the voice gateway to retry other dial-peers.

There's 3 dial-peer which matches this e164 number.
1)Dial-peer 100 goes CUCM (longest match, most specific)
2)Dial-peer 200 goes CUCM (longest match, most specific)
3)Dial-peer 300 goes to PSTN (the destination-pattern is .T)

When dial-peer 100 and 200 "fails", the voice gateway will dial-out to PSTN
via dial-peer 300. Once again, PSTN route back to the customer VG. This
causes a routing loop and it can fills up all the available E1 channels
quickly.


*Just wondering if anyone encounter the following issue and have a
explanation to it? Just the engineering side of me want to get down to the
root cause. *

The CUCM have "stop routing on unallocated number" turns off (false). Just
in case it matters.

I tried to google around but can't seems to find any article that talks
about
1) dial-peer behaviors (on voice gateway side) - on what error code will
cisco voice gateway retry other dial-peers?
2) why CUCM returns error code 27?

It's a managed service system so I'm unable to do a deep dive
troubleshooting.

The current workaround introduced is to create a dial-peer 250 with a
higher preference that matches the DID range and block it.

This means that incoming dialed number will match to 4 dial-peers (100,
200, 250 and 300)

After failing on 100 and 200, the call gets block on dial-peer 250.

-- 
Regards,
Ki Wi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/cisco-voip/attachments/20170814/09024b49/attachment.html>


More information about the cisco-voip mailing list