[asterisk-ss7] chan_ss7 - T22 timeout (No 'circuitgroup resetacknowledge' from peer)

[Jakub Klausa]

On Sat, Jul 19, 2008 at 07:35:43PM +0200, Pawel Ratajewski (Forweb) wrote:

=> with some telcos it works fine:
=> 
=> ss7 block 10 1
=> Sending Blocking message to peer
=> Jul 16 12:54:46 NOTICE[10270]: l4isup.c:3789 do_group_circuit_block_unblock:
=> Sending CIRCUIT GROUP BLOCKING, cic=10, mask=0x00000001.
=> Jul 16 12:54:46 NOTICE[5425]: l4isup.c:3270 process_cga: Process CGA,
=> cic=10, range=32

Depends on what you call 'fine'. You've sent a block for a single CIC,
you've received an CGA for 32 CICs. If that's 'fine' by your standards, then
it's fine allright. But generally speaking it's not. 

=> with other does not:
=> 
=> ss7 block 10 1
=> Sending Blocking message to peer
=> [Jul 16 13:22:35] NOTICE[13152]: l4isup.c:3793
=> do_group_circuit_block_unblock: Sending CIRCUIT GROUP BLOCKING, cic=10,
=> mask=0x00000001.
=> [Jul 16 13:22:35] DEBUG[13152]: l4isup.c:389 mtp_enqueue_isup_packet: Queue
=> packet CIC=10, len=16, linkset='exatel', link='l1', slinkset='exatel',
=> slink='l1'
=> [Jul 16 13:22:35] DEBUG[12919]: mtp.c:1867 mtp_thread_main: Queue MSU,
=> lsi=0, last_send_ix=0, linkset=exatel, m->link=l1
=> [Jul 16 13:22:35] DEBUG[12919]: mtp.c:1575 mtp2_fill_zaptel_buf: Sending
=> buffer to zaptel len=20, on link 'l1' bsn=35, fsn=51.
=> 
=> [Jul 16 13:23:05] WARNING[12919]: l4isup.c:1122 t18_timeout: T18 timeout (No
=> "circuit group blocking acknowledge" from peer) CIC=10.

If you have a single span in the linkset l1, or you have less then 42 CICs
assigned to that linkset, then everything is allright here. Don't be fooled
bu the do_group_circuit_block_unblock message - the real mask sent is not
what it says it is. Connect your protocol analyzer to that link, monitor it
for a while. See what really gets sent there. I think it's a known bug in
chan_ss7 it sends the CGB/CGUs wrong (the range is always 32 there).

That's why you see the CGA with range=32 in the first example, despite
trying to send ithe CGB with range=1. 

The problem should have come out during the interconnect testing procedure
though...

-- 
Jakub Klausa | j.klausa at ss7.pl | http://www.ss7.pl/ | http://www.ngpbx.pl/
Dane rejestrowe -> http://kontakt.ss7.pl 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
Url : http://lists.digium.com/pipermail/asterisk-ss7/attachments/20080719/20c4327e/attachment.pgp