ISDN problem after office move 1

[richardmis]

Hi guys

We are a UK based company.

We have Cisco 2610 ISDN Router that has worked fine for the last five years. We have now moved offices, and it is unhappy. We use the router to dial out to customer sites. The config has not changed since the move, could not see any reason why the config would need to change. The ISDN lines are up and running, had BT on site yesterday to check the lines. There seems to be a connection between the ISDN terminals and the Router. If I pull a cable out the light on the ISDN terminal starts flashing, if I plug the cable back in it stops flashing. Now if I try to connect to a remote site it attempts to dial the correct number i get

ISDN-6-LAYER2DOWN: Layer 2 for Interface BR1/3, TEI 255 changed to down

I don't want to post all the debug info at this stage, if someone can point me in the right direction. If not I will post the debug info.

Thanks

Richard

 

[ISDNman]

Was the old line or a multipoint line?

Most equipment can support either. However if you are configured multipoint you will not work on a point-to-point line.

Point-to-point configured equipment will work a multipoint line, so the Telco's really should always go with Multipoint, but if they made things to easy I think the lose their membership in the secret Telco society .

The other possibility is that you are marginal (distance wise or termination resistor-wise) between the NTBA/Network Termination 1 . I would try testing with a 1 meter cable from the BT's NTBA and your Cisco and see if that makes a difference.

Best of luck.

 

[rtfmdude]

Hi richardmis! If I remember correctly, the TEI message you posted is basically the router trying to reset the BRI at the q921/OSI layer 2 level.

It's been forever since I've had to troubleshoot a basic-net3 circuit, so, I can't remember 100% what a TEI of 255 means. I'm pretty sure it means the BRI interface wasn't able to get one from BT's local ISDN switch when the interface attempted to dial out.

lemme see, since euro isdn (net3) does the whole power-saving thing and tears down layer 2 then layer 1 after an idle period, I think the order of operation is something like this:

1 - interface is idle, with a live isdn circuit. Should be in a low/no power state, I believe.

2 - dialout attempt is made, which starts a chain of events:

2a- 'toggle' by the router's BRI interface at layer 1, an isdn equivalent of flipping on a light switch. This should bring the circuit up to allow layer 2 to negotiate.

2b- the router requests a TEI (terminal endpoint identifier) from the local BT switch. This is normally a dynamic value and doesn't require that you specify anything (like dhcp). BT should assign a TEI to you, I think a value of 64-126 would be appropriate (that's the dynamic TEI range). The one thing that I think would be relevant would be to call up BT and make sure they don't expect you to have a static TEI configured rather than a dynamic (the TEI message you posted seems to indicate this would be a likely culprit). Anyway, obtaining a TEI is a necessary part of establishing a communication path with the local BT switch, after that you should be able to place calls.

I looked for some internet examples of the TEI negotiation for a euro circuit, unfortunately all I could find was in google's cache:

http://209.85.207.104/search?q=cach...+1&hl=en&ct=clnk&cd=10&gl=us&client=firefox-a

the logs there are basically the output of 'debug isdn q921' and 'debug isdn q931' - grab that if you still have problems after talking to BT. I'll try to annotate the debugs in-line to break out the steps as I attempted to outline them above, look for [RTFM]:


CE-3.LAB#isdn test call in bri0 019161

[RTFM] - ok, this guy uses the 'isdn test call' command to manually kick off a call rather than using dial on demand routing. it's a useful way to do some basic tests. This should 'wake' the circuit up.

CE-3.LAB#

*Oct 28 17:06:12.767 CEST: ISDN BR0 **ERROR**: handle_l2d_srq_mail: Layer 1 inactive
[RTFM] - notice layer 1 was inactive previously on this euro circuit. power-saving measure. kyoto-zealots

*Oct 28 17:06:12.967 CEST: ISDN BR0 Q921: User TX -> IDREQ ri=1463 ai=127

[RTFM] - ok, there wasn't any output saying that layer 1 was brought up, but it's assumed here since the router is requesting a TEI from the local switch. It's sending an ID REQUEST with a current value of 127, which is the 'broadcast' dynamic TEI value. Kinda like sending a DHCP DISCOVER with a source address of 0.0.0.0

*Oct 28 17:06:12.983 CEST: ISDN BR0 Q921: User RX <- IDASSN ri=1463 ai=120
[RTFM] - local telco switch responds and offers us a TEI value of 120.

*Oct 28 17:06:12.987 CEST: ISDN BR0 Q921: L2_EstablishDataLink: sending SABME
[RTFM] - router saying 'ok, it's time to acknowledge/setup communications utilizing TEI value 120

*Oct 28 17:06:12.987 CEST: ISDN BR0 Q921: User TX -> SABMEp sapi=0 tei=120
[RTFM] - router sends a SetAsynchronousBalancedModeExtended (SABME) message for TEI 120, kinda like a DHCP ACK from the client to the server after the server hands down an ip address

*Oct 28 17:06:12.999 CEST: ISDN BR0 Q921: User RX <- UAf sapi=0 tei=120
[RTFM] - switch sends back an UnAssignedframe (UAf), acknowledging the assignment. At this point, the TEI is considered 'multiple frames established' (the ONLY good state for passing calls - check 'show isdn status' to see this).

*Oct 28 17:06:13.003 CEST: %ISDN-6-LAYER2UP: Layer 2 for Interface BR0, TEI 120 changed to up
[RTFM] - notice the router says the interface is up now.

*Oct 28 17:06:13.003 CEST: ISDN BR0 Q921: User TX -> INFO sapi=0 tei=120, ns=0 nr=0
[RTFM] - this layer 2 (q921) INFO message corresponds to the layer 3 (q931) SETUP message below, which contains all the called/calling party numbers/etc to get the call to your remotes established. This is probably further than you're getting so I'll stop here.

*Oct 28 17:06:13.003 CEST: ISDN BR0 Q931: SETUP pd = 8 callref = 0x04

I would check ISDNman's suggestion on the cable issue - a marginal layer 1 connection casts doubt on all the other stuff that rides on top of it. (expect coffee to spill in cupholder because you're going 100kph in a Smart car with 3 flat tires kinda thing)

post the output from a debug isdn q921/q931 when you're trying to dial out if you're still having issues!