[Home]

Summary:ASTERISK-07753: NFAS primary D-channel always being maked as ACTIVE on startup
Reporter:Larry Butler (larrybutler)Labels:
Date Opened:2006-09-15 15:19:48Date Closed:2011-06-07 14:02:51
Priority:MajorRegression?No
Status:Closed/CompleteComponents:Core/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) larrysheadache.txt
( 1) nfas_fix.diff
Description:I have an ISDN NFAS group configured in zapata.conf with primary and seconday D channels.  If the carrier fails over signaling to Secondary and asterisk is restarted, the primary D channel is alway marked as ACTIVE and secondary marked STANDBY.  As a result signaling is split (Inbound is being received on the secondary and outbound is being sent on primary) and the call is terminated.

zapata.conf
...
; NFAS Configuration
[trunkgroups]
trunkgroup => 1,24,48

spanmap => 1,1,0
spanmap => 2,1,1
spanmap => 3,1,2
spanmap => 4,1,3
...





****** ADDITIONAL INFORMATION ******

Not sure if this is happening when the DS1 with Primary D comes out of alarm as well.
Comments:By: Larry Butler (larrybutler) 2006-09-15 15:35:19

Looking for a workaround, I find that if I run /sbin/ztcfg the primary is kicked over to active on the carrier side and all is well.  I can add this to the rc script but am unsure if this is a viable work around.

By: cmaj (cmaj) 2006-09-17 18:57:58

Check out bug 3450.  I don't have a test environment for it right now, but I would be happy to help update to current CVS if you can test.  I think a combination of that patch with a little new code will produce the functionality you are looking for.  It's been a while since I've looked at it, though.

By: Larry Butler (larrybutler) 2006-09-18 10:05:43

Thanks!  It will take me a couple of days to get an NFAS setup in our Dev environment, but I will be happy to assist and test.

By: Larry Butler (larrybutler) 2006-09-21 15:01:38

I would like to help out on 0003450 because I think the functionality will be useful.  But I believe this issue should also be addressed in main trunk.

If someone could guide me on where to start looking I could address and contribute.  Any ideas?

By: Serge Vecher (serge-v) 2006-09-22 08:25:52

cmaj, larrybutler: if you guys have an updated patch for 3450, please find me on irc #asterisk-bugs or drop me an email at svecher at gmail dot com I will reopen that bug. Thanks!

By: cmaj (cmaj) 2006-09-22 11:06:00

serge-v: Please do re-open 3450.  I've almost completed updating those patches against trunk.

larrybutler: Testing those patches will be critical, along with full PRI dumps, preferably on both CPE & NET sides.

By: Larry Butler (larrybutler) 2006-09-22 11:10:12

You got it.  It's taking me longer than expected to get NFAS setup our dev environment.  The current order completion date is set for 10/03.

By: cmaj (cmaj) 2006-09-22 14:02:54

New patches are in 3450.  After the asterisk restart, a "pri service disable channel 24" on the CLI should disable the D-channel on the first PRI and trigger the needed messages for asterisk to fail over to the secondary D-channel.



By: jmls (jmls) 2006-11-01 12:55:07.000-0600

ping. where are we with this ?

By: Matthew Fredrickson (mattf) 2006-11-01 13:23:34.000-0600

We're waiting on feedback on a patch that I wrote to fix the problem.

By: Larry Butler (larrybutler) 2006-11-01 14:05:24.000-0600

Sorry guys.  NFAS 94x2 has been configured in our Dev environment and the patch is applied.  Just waiting for a resource availability from the NET side to test.

By: Matthew Fredrickson (mattf) 2006-11-10 11:29:38.000-0600

Hey, any feedback on this?  I took the time to write a patch for it, and I'd appreciate knowing that it didn't go to waste.  Thanks!

By: Serge Vecher (serge-v) 2006-11-15 11:13:52.000-0600

matt: is the patch you are talking about attached to this or some other bug?

By: Matthew Fredrickson (mattf) 2006-11-15 11:56:16.000-0600

Sorry, no it wasn't.  I had talked to him directly at Astricon, and directly sent the patch to him. I just attached it to the bugnote.

By: Matthew Fredrickson (mattf) 2006-11-22 13:11:43.000-0600

Suspended due to lack of response.