Knowledge Share – ISDN Se0/2/1:23 **ERROR**: L3_BadPeerMsg – Part II

Analysis
Nov 10, 20082 mins

Knowledge Share – ISDN Se0/2/1:23 **ERROR**: L3_BadPeerMsg - Part II

Well, that was an interesting experience, got a few interesting replies but most people would still go to TAC, which makes sense, and I would have done the same if it was not resolved the following day, but it did, here is how. As it came to pass, at some point I was certain the problem only appears when making more then one call to a specific number, out of the four that rings this line. That number was ported over the day before from an analog circuit to this PRI. My senses were that perhaps it is just forwarded somehow but there are still limitations on the amount of concurrent calls. We contacted the phone company and tried to verify what is going on with this number from their end; according to them the number was moved and not forwarded. There is no reason to this behavior from their perspective. Still the problem was there, and had no systematic appearance, so we needed the phone company’s help. Started to troubleshoot it with them and it appeared as if on the failed calls something is not happening right. The setup message was getting a disconnect from our side. The router was rejecting the calls for some reason. A typical ISDN call flow would look like the following: [img]https://sites.google.com/site/avnerizhar/Home/pri-flow.gif[/img] In our failed calls, the setup and the facility were coming in, but the instead of call proceeding we were getting a ‘disconnect’, with the reason code saying “Invalid call reference value”. That was not making any sense and the router was not supposed to reject the calls. Perhaps the facility-ie messages are causing it, we don’t have to have them, lets disable them I told the phone company guys, but they said it is what allows us to have a calling name under national isdn II signaling. National ISDN??? But my side is set to DMS100, like it was on the old site … Well, that was the problem, once changed it all started to work like it should and we started to see names in the ‘debug isdn q931’. That’s it; it all came back to the basics, hoping to save others from going through the path I did, at least now when I Google it this solution comes up, later …