h.323 call preservation don't work well. Call disconnections

pavelsh_ucs · ‎04-20-2010

Hello!

I had a working solution for our remote office. All worked well except case when our link to provider goes down.

If the link goes down then all active calls drop out, and phones began register on SRST. Providers problems were not so frequently but many people were very angry in such case.

So I decided to setup call preservations. I've setup a necessary parameter on cucm service parameters and setup call preservation with this script on our cisco gateway:

conf t
voice iec syslog

gateway
timer media-inactive 5

voice service voip
h323
call preserve
no h225 timeout keepalive
end

wr mem

All ok. But now I have a big problem. In random time calls were disconnected by call preservations.

For example:

Apr 19 13:43:01.513: %CCH323-6-CALL_PRESERVED: cch323_h225_handle_conn_loss: H.323 call preserved due to socket closure or error, Call Id = 7112, fd = 12
Apr 19 13:43:01.513: %VOICE_IEC-3-GW: H323: Internal Error (Socket error): IEC=1.1.186.5.7.6 on callID 22825

Apr 20 07:46:54.936: %CCH323-6-CALL_PRESERVED: cch323_h225_handle_conn_loss: H.323 call preserved due to socket closure or error, Call Id = 9673, fd = 8
Apr 20 07:46:54.936: %VOICE_IEC-3-GW: H323: Internal Error (Socket error): IEC=1.1.186.5.7.6 on callID 24437

Apr 20 09:47:30.705: %CCH323-6-CALL_PRESERVED: cch323_h225_handle_conn_loss: H.323 call preserved due to socket closure or error, Call Id = 7021, fd = 3
Apr 20 09:47:30.705: %VOICE_IEC-3-GW: H323: Internal Error (Socket error): IEC=1.1.186.5.7.6 on callID 25344 GUID=5B441FBA4BD211DF898E002545FC7ED0

Apr 20 13:48:21.452: %CCH323-6-CALL_PRESERVED: cch323_h225_handle_conn_loss: H.323 call preserved due to socket closure or error, Call Id = 5834, fd = 28
Apr 20 13:48:21.452: %VOICE_IEC-3-GW: H323: Internal Error (Socket error): IEC=1.1.186.5.7.6 on callID 27300 GUID=80EE3751DA11E1BCB50EC7010A0A257C

So internal socket error hung down live call. All other calls working well in this time. In one day from 2 to 5 calls.

How I can fix this? I see only one fix - disable call reservation.

sh ver
Cisco IOS Software, 3800 Software (C3825-ADVIPSERVICESK9-M), Version 12.4(24)T1, RELEASE SOFTWARE (fc3)
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 1986-2009 by Cisco Systems, Inc.
Compiled Fri 19-Jun-09 18:40 by prod_rel_team

gogasca · ‎04-23-2010

Logs point to a lot of things. We better have the following debugs

 
Could you provide me the following debug information for calls that are
having this issue
 
config t
service timestamps debug datetime local msec
service timestamps log datetime local msec
service sequence
no logging console
no logging monitor
no logging rate-limit
no logging queue-limit
logging buffer 3000000 debug
voice iec syslog
 
debug isdn q931
debug voip ccapi inout
debug h225 asn1
debug h245 asn1
debug ip tcp transaction
debug voip vtsp default
debug voip vtsp session
 
 
After turning on the debugs, issue the command "clear log". Then reproduce
the issue. After that, dump the contents of the log using:
term len 0
show log

Whats your IOS version?
Whats your CUCM version?
Can you add:
h225 timeout tcp establish 0
to voice class or dial-peer and see if that makes any difference?
So we would like to know if by disabling call preservation 
you still the same number of drops.?
All calls are dropped inmediatly after error is generated?

Do you have media inactivity detection configured? 
Check below:

*RTP and RTCP Inactivity Detection Configuration Example *

/The following configuration example can be used to enable RTP and RTCP
inactivity detection for dial peers. Note that for call preservation VAD
must be set to off (no vad command):/

dial-peer voice 10 voip
no vad
gateway
timer receive-rtcp 4
ip rtcp report-interval 60

*Bidirectional Silence Detection Enable Example *
/The following configuration example enables bidirectional silence
detection: /
gateway
timer media-inactive 5
ip rtcp report interval

Let us know