Hello, - I have tried several NMCs (including HiperNMC) none of them had anything customised other than the snmp community strings. BTW, I do have chassis awareness enabled on the HARC. - I had suspected in the past an NI-2 issue but how do you explain that when I "manually" busy out a channel, the switch skips the channel correctly ? It is only when the DSP locks into LOOS that I get a fast busy. - One thing I forgot to mention is that some dial-in users (very few) get a Radius Attribute of Tunnel-Server-Endpoint. In other words, the HiperArc is acting as a LAC for an L2TP tunnel. This is probably unrelated but at this point I am suspecting anything... - One of 3com tech support did acknowledge the issue a while ago, he sent me a special HARC code (5.0.81) but it did not fix the problem. He suggested I go back to 4.2.73 I'm just curious on what exactly is causing this behaviour Donald
Also sprach Donald Baud
I find it hard to believe that much of anything on the Arc could be causing channels to be going to LocalOutOfService. Do you have any autoresponse stuff set up on the NMC? That's the only thing I can think of there.
One other observation: - Once a channel is stuck in LocalOutOfService, any further call beyond that channel on the PRI gives a fast busy signal
Sounds like you're running NI-2 translation on it. NI-2 doesn't have service message, so the switch doesn't know the channel is LocalOutOfService, so it keeps sending calls down that channel. The DSP does the only thing it knows to do and rejects the call resulting in the fast busy. Switch to a custom-5ESS or custom-DMS100 (or whatever switch type your telco has) and you should be able to get service messages and get the behaviour you're expecting here (LocalOutOfService channels get skipped) -- Jeff McAdams Email: jeffm@iglou.com Head Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
---- Original message ---------------- I am trying to move to the 5.x.x Hiper Arc code branch but for some reason The DSP's running in the same chassis start to show Local Out Of Service on one or two channels after a few days of service. The chassis is configured as a regular IP only dial-in service.
Here are my observations: - The issue happens with all recent DSP codes (up to 2.1.9) - The issue happens with all Hiper ARC 5.x.x (up to 5.1.99-8 including 5.0.x) - The issue does NOT happen with Hiper Arc 4.2.x (up to 4.2.78) - The issue happened with DMS100 and 5ESS switches connected to the DSPs - I am running PRI with no NFAS (i.e 23 channels + D channel) - The issue does NOT affect the PRI T1/E1 Quad Modems - I tried NMC and HiperNMC without any difference - In all my tests I was running ospf - I have tried 64Megs and 128Megs HiperArc cards - I made sure to enable ppp offloading - the issue happens with chassis of 2 DSP's up to 10 DSPs
When the problem happens I get the Service state 8 on TCM. If I busy/Unbusy the DSP card the channels lock out and show a "Remote Out Of Service" message and the PRI channels can nolonger be busied/unbusied until I reboot the DSP card. I usually also reboot the HiperArc just in case.
I refuse to think that this is related to telco because I have been able to repeat the issue with two different PRI providers with two different switches. And the issue happens only with HiperArc 5.x.x (never with 4.2.x)
- To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.