[Home]

Summary:ASTERISK-08305: [patch] When using L option on Dial, instead of warning asterisk disconnects the call
Reporter:Richard Wilkinson (rickead2000)Labels:
Date Opened:2006-12-07 14:38:46.000-0600Date Closed:2006-12-12 22:40:12.000-0600
Priority:MajorRegression?No
Status:Closed/CompleteComponents:Applications/app_dial
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) 8541.patch
Description:When using the L option in a Dial command to limit the length of a call and providing values to provide a warning at, asterisk disconnects the call at the first warning value.

e.g.

DIAL (SIP/1234,30,L(180000:60000:30000))

Will disconnect the call when there are 60s to go, instead of providing a warning.

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

Dec  7 20:26:56 VERBOSE[22167] logger.c: AGI Rx << EXEC DIAL SIP/1234|30|L(180000:60000:30000) ""
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- AGI Script Executing Application: (DIAL) Options: (SIP/1234|30|L(180000:60000:30000))
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- Limit Data for this call:
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - timelimit     = 180000
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - play_warning  = 60000
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - play_to_caller= yes
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - play_to_callee= no
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - warning_freq  = 30000
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - start_sound   = UNDEF
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - warning_sound = timeleft
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- - end_sound     = UNDEF
Dec  7 20:26:56 DEBUG[22167] chan_sip.c: Setting NAT on RTP to 524288
Dec  7 20:26:56 DEBUG[22167] chan_sip.c: Outgoing Call for 1234
Dec  7 20:26:56 VERBOSE[22167] logger.c:     -- Called 1234
Dec  7 20:26:56 DEBUG[20929] chan_sip.c: Acked pending invite 102
Dec  7 20:26:56 DEBUG[20929] chan_sip.c: Stopping retransmission on '2f355fc04814f1a4286b85c30e4c2469@lvdx.com' of Request 102: Match Found
Dec  7 20:26:56 DEBUG[20929] chan_sip.c: (Provisional) Stopping retransmission (but retaining packet) on '2f355fc04814f1a4286b85c30e4c2469@domain.com' Request 103: Found
Dec  7 20:27:03 DEBUG[20929] chan_sip.c: (Provisional) Stopping retransmission (but retaining packet) on '2f355fc04814f1a4286b85c30e4c2469@domain.com' Request 103: Found
Dec  7 20:27:03 DEBUG[20918] channel.c: Avoiding initial deadlock for 'SIP/1234-00652dd0'
Dec  7 20:27:03 VERBOSE[22167] logger.c:     -- SIP/1234-00652dd0 is ringing
Dec  7 20:27:03 DEBUG[22167] channel.c: Driver for channel 'SIP/70-aca16320' does not support indication 3, emulating it
Dec  7 20:29:04 DEBUG[22167] channel.c: Nobody there, continuing...
Dec  7 20:29:04 DEBUG[22167] channel.c: Bridge stops bridging channels SIP/70-aca16320 and SIP/1234-00652dd0
Dec  7 20:29:04 DEBUG[22167] chan_sip.c: update_call_counter(1234) - decrement call limit counter
Dec  7 20:29:04 DEBUG[22167] app_dial.c: Exiting with DIALSTATUS=ANSWER.
Comments:By: Serge Vecher (serge-v) 2006-12-07 14:44:27.000-0600

I think this was recently fixed, can you please check out 1.2 from svn and check?
By: Anthony LaMantia (alamantia) 2006-12-07 16:11:57.000-0600

yes,  I also recall a patch resolving this issue or one much like it. if you can please test this out with the latest 1.2svn branch and see if the problem is still happening on this build.
By: Richard Wilkinson (rickead2000) 2006-12-10 16:16:35.000-0600

Latest 1.2svn branch is still broken!
By: Tony Howat (tonyh) 2006-12-11 08:52:49.000-0600

This bug applies to zaptel calls as well as SIP.

Looks like ASTERISK-8155 was the same thing.

I can also confirm that latest 1.2svn is still broken.


By: Serge Vecher (serge-v) 2006-12-11 10:41:41.000-0600

murf: can you please investigate these claims?
By: Serge Vecher (serge-v) 2006-12-11 10:42:55.000-0600

tonyh, rickhead2000: please specify which revisions you've observed these in 1.2 branch... thanks.
By: Tony Howat (tonyh) 2006-12-11 12:17:39.000-0600

I tried full 1.2 SVN checkout today, r48233 of channel.c.

My issue is with a normal zaptel call, and it's still exhibiting the same behaviour - ie. hanging the call up instead of warning.

Thanks for looking at this...
By: Tony Howat (tonyh) 2006-12-11 12:33:34.000-0600

-- Executing Dial("Zap/1-1", "Zap/g1/07xxxxxx43||L(60000:30000:5000)") in new stack
   -- Limit Data for this call:
   -- - timelimit     = 60000
   -- - play_warning  = 30000
   -- - play_to_caller= yes
   -- - play_to_callee= no
   -- - warning_freq  = 5000
   -- - start_sound   = UNDEF
   -- - warning_sound = xtest/warn
   -- - end_sound     = xtest/terminate
   -- Requested transfer capability: 0x00 - SPEECH
   -- Called g1/07xxxxxxx43
   -- Zap/2-1 is proceeding passing it to Zap/1-1
   -- Zap/2-1 is ringing
   -- Zap/2-1 answered Zap/1-1

(after 30s)

   -- Hungup 'Zap/2-1'
 == Spawn extension (extdialbridge, 1, 107) exited non-zero on 'Zap/1-1'
      > cdr_odbc: Query Successful!
   -- Hungup 'Zap/1-1'


Note that samples etc are there, this stuff worked previously under 1.2.7 for me.


By: cmaj (cmaj) 2006-12-11 20:15:48.000-0600

Same problem in 1.4 beta 3.
By: Steve Murphy (murf) 2006-12-12 07:49:14.000-0600

Amazing! I can reproduce the problem! How on EARTH did I not see this problem with  the previous bug?

OK, enough fooling around. I'll analyze the code that's there, and rewrite it if I have to, to make it work. And if this problem extends into 1.4 and up, I'll get it there, too.

By: Tony Howat (tonyh) 2006-12-12 08:27:33.000-0600

Excellent stuff murf... thanks!
By: Steve Murphy (murf) 2006-12-12 13:59:58.000-0600

I have uploaded small patch. This is a matter of one bug's fix creating another; the fix I made previously for 8386 was valid, and fixed the problem, however, a subsequent bug fix on a DTMF issue broke the code. Hopefully, they are now both happily working.

Try out my patch, and see if works. If it does, I'll commit it to 1.2.
By: Steve Murphy (murf) 2006-12-12 14:36:39.000-0600

I just did a 1.4 stock latest svn build and install, and tested the L(x:y:z) in dial, and it works fine. cmaj, you need to tell me EXACTLY what you did to 1.4, when it fails, or I won't be able to fix it.
By: Richard Wilkinson (rickead2000) 2006-12-12 15:26:08.000-0600

murf - I can confirm that patch resolves the issue when patched against the latest 1.2svn - nice work!
By: Steve Murphy (murf) 2006-12-12 22:35:12.000-0600

working patch applied to 1.2 with r.48434. and blocked from 1.4 via r.48435