Channels sticking open and costing money on voip

[davesdatasystems]

The issue is back, and i need to figure this out, hopefully someone has found a fix. I dont know what other settings i can change to prevent this. Two dialers Different dial plans. Different ISPS different VOIP providers.
I would like to know if some of the people who had this issue in the past, if the issue went away with vicidial7 (asterisk 11)


chan_sip.c:4058 __sip_autodestruct: Autodestruct on dialog '236b4eb2638f37063d4c8a8450ad6d21@108.9.141.70:5060' with owner SIP/ IPLINK1-00005d95 in place (Method: BYE). Rescheduling destruction for 10000 ms

At times getting the RTP at 61 seconds error as well.

when i did sip show channels. This is what i seen
38.102.250.60 19162244268 35962ac95e65c81 0x4 (ulaw) No Tx: ACK Xcast
38.102.250.60 15738936502 3b8c94bc44dbe90 0x4 (ulaw) No Tx: ACK Xcast
172.16.0.101 (None) VTuE_lyvUgd0asE 0x0 (nothing) No Rx: SUBSCRIBE <guest>
XXXX.XXXX.XXX.XXX 023 68656bb52e0b16f 0x4 (ulaw) No Tx: ACK 023
38.102.250.60 15629482077 3c02c8980deb47c 0x4 (ulaw) No Init: INVITE Xcast
38.102.250.60 13105746160 186b125178186df 0x4 (ulaw) No Init: INVITE Xcast
XXXX.XXXX.XXX.XXX 801 414ecdde7174f85 0x4 (ulaw) No Tx: ACK 801
38.102.250.60 16267917440 7a50e3c146ced94 0x8 (alaw) No Init: INVITE Xcast
38.102.250.60 16268404926 7e07d6d34625331 0x4 (ulaw) No Init: INVITE Xcast
173.78.237.79 040 67ca12bc516bec0 0x4 (ulaw) No Tx: ACK 040
38.102.250.60 17046980033 1c41b4bc3299ac8 0x4 (ulaw) No Init: INVITE Xcast
24.73.220.146 030 071749be4c09bd7 0x4 (ulaw) No Tx: ACK 030
38.102.250.60 18314234650 5ae30fdb2ab02b4 0x4 (ulaw) No Tx: ACK Xcast
38.102.250.60 14102727060 49dadc0c7f82156 0x4 (ulaw) No Tx: ACK Xcast
38.102.250.60 14409344891 05ca11c8159f4d2 0x4 (ulaw) No Init: INVITE Xcast
38.102.250.60 18022476158 1c94a138475d3b3 0x4 (ulaw) No Tx: ACK Xcast
38.102.250.60 13103769189 5ce803a2395045a 0x4 (ulaw) No Init: INVITE Xcast
38.102.250.60 16239744667 51f2527d2d37d86 0x4 (ulaw) No Init: INVITE Xcast
38.102.250.60 13232251701 184e51ef1a376cd 0x4 (ulaw) No Init: INVITE Xcast
38.102.250.60 15862944422 3d02468c15e3cdc 0x4 (ulaw) No Tx: ACK Xcast
38.102.250.60 14017229129 4d64dc0076dba81 0x4 (ulaw) No Init: INVITE Xcast
24.73.220.146 029 184daf0a63597ec 0x4 (ulaw) No Tx: ACK 029
172.16.0.102 004 4bd7cbdf5da1fa8 0x4 (ulaw) No Tx: ACK 004
38.102.250.60 15624204464 691a5d3f178e30a 0x4 (ulaw) No Init: INVITE Xcast
173.78.237.79 039 018cbd1c6a0a0a8 0x4 (ulaw) No Tx: ACK 039
24.73.220.146 024 24228510567f420 0x4 (ulaw) No Tx: ACK 024
38.102.250.60 19017590152 306db1c7466153b 0x4 (ulaw) No Tx: ACK Xcast
38.102.250.60 15629042681 58b3d51639f8f7e 0x4 (ulaw) No Init: INVITE Xcast
38.102.250.60 18054960141 690ef71d2deece7 0x4 (ulaw) No Tx: ACK Xcast

Fixes i have attempted. I added () to the end of the dial plan after hangup, and it seemed to slow it down. The systems have been restarted every night. The big issue is it is causes a lot of extra money to be spent on VOIP, i really need to get some help on this one. The channels are stuck open even when they do not go to agents, so please advice on how to shut down these channels or how to fix the issue. The ISP speeds are also fine. ONe of the systems having this issue only has 5 agents and is 35/35 MBPS. I also made sure that Asterisk version was correct. in vici. I am at a loss, i know others have had this issue. so hopefully i can get a answer that will work for me this time.

Dialer One DIal plan

[IPLINK1]
type=peer
host=XXX.XXX.XXX.XXXX (carrier ip is correct, but do not like posting IPs)
canreinvite=no
qualify=yes
disallow=all
allow=ulaw,alaw
dtmfmode=rfc2833
context=trunkinbound

exten => _9COC1NXXNXXXXXX,1,AGI(agi://127.0.0.1:4577/call_log)
exten => _9COC1NXXNXXXXXX,n,Dial(SIP/IPLINK1/${EXTEN:4},,tTor)
exten => _9COC1NXXNXXXXXX,n,Hangup()

Dialer 2 Dail plan

[Xcast]
insecure=port,invite
secret=no
qualify=yes
host=(carrier ip is correct, but do not like posting IPs)

disallow=all
allow=ulaw,alaw,g729
dtmfmode=rfc2833
nat=no
canreinvite=no
context=from-internal
type=peer


insecure=port,invite
secret=no
qualify=yes
host=(carrier ip is correct, but do not like posting IPs)

disallow=all
allow=ulaw,alaw,g729
nat=no
canreinvite=no
context=from-trunk
dtmfmode=rfc2833
type=peer

If you need any more information just ask here or in private, This has to get resolved.

Also i know this is in asterisk, has anyone had this issue, then upgraded to asterisk 11 and the issue go away?

[rrb555]

I had the same issue. What is ur max trunk and max call per second value?

I set my servers to reboot nightly which kinda help. I always thought that this is an asterisk issue but i believe it is not

[davesdatasystems]

I had the same issue. What is ur max trunk and max call per second value?

I set my servers to reboot nightly which kinda help. I always thought that this is an asterisk issue but i believe it is not

I think it might of been. Was you running 1.8?

I just said screw it, and upgraded them to vicibox 7 with asterisk 11, I will know in a week or so if it is still a issue, because it comes on about a week after install. Kind of sucks, because it is hard to trouble shoot after a reinstall

Max trunk is set a 200
max call at 20

I had the issue, did the nightly restarts, but once the issue hit, it would not go away, and the autodestruct would go completely crazy and i knew there was a issue. Just for the future, if i may need to know, what ever steps did you take to resolve it.

But if everything is good in a week, i will know it was a some kind of bug in 1.8.

[rrb555]

You shouldnt set ur max trunk like that. It should be around 95 based from
Mu settings. And call is somewhat same value. I change mine to that value and it works now.

Yep i am running 1.8. I had same thinking just like u do before but when i read a thread here and change those settings and it just works

[mflorell]

If you are using SIP your Max Calls per Second should be much lower than 20. I would try setting it to 4.

[williamconley]

I've had clients who had to set their (press one dialer) max calls per second to over 100 to maintain the dial level they needed. Depends a LOT on the carrier. But I've never had a client with "stuck" calls even with high max calls per second unless they were simply overloading their server.

I will say that the poweredge 1850 is a wee bit old and 4 cores total is a wee bit small. But if you're staying under 200 calls on the server you should not really be in overload.

However: You did not post any aggregate data on the occurrences. I'd suggest you begin tallying information about what these calls and "occurrences" may have in common. You may find a report getting run that overloads sql and stops an agi script from completing on a bunch of calls at regular intervals or the same time of day. If you're lucky.

If you're even luckier, asterisk 11 will solve it. But a bigger server may also be your solution. Eight cores is better than 4.

[davesdatasystems]

Went with a more powerful system and went to 11, have not seen the issue, also told people to keep the max call per second at 4 just to be on the safe side. I have not had any issues yet, after two more weeks, i will visit this topic again in about a week and if all is resolved i will go through all the steps i took.