×

Warning message

  • Cisco Support Forums is in Read Only mode while the site is being migrated.
  • Cisco Support Forums is in Read Only mode while the site is being migrated.

Caller doesn’t get busy notification if called phone is busy, instead user get “congestion” notification

Answered Question
Feb 1st, 2014
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Dear All,


Recently we have replaced legacy SIP server with Cisco SME after that users observed that, Caller doesn’t get busy notification if called phone is busy, instead they get “congestion” notification.


Below is the Call flow:


Alcatel Callmanager 1---SIP Trunk----Cisco SME----SIP Trunk----Alcatel Callmanager 2


I have analyzed SIP logs and found that Cisco SME doesn’t relay “Buys here, 486 message” back to Originator callmanager, in above scenario when someone call from Alcatel Callmanager 1’ phone to Alcatel Callmanager 2’ phone (which is Busy, already on call). In this case Alcatel Callmanager 2 sends “Busy here, 486 message” to Cisco SME and Cisco SME sends ACK message back to it but Cisco SME doesn’t relay this message back to Alcatel Callmanager 1, and after few seconds Alcatel Callmanager 1 sends “CANCEL” message to Cisco SME and then Cisco SME ACK to it and sends “Cancelled request” to Alcatel Callmanager 1 .


I'm attaching the Cisco SME SDI logs and below is the info..


From: 10.205.39.8 (Alcatel Call manager 1 IP address)

Caller Phone: 6206698

To: 10.205.228.9 (Alcatel Call manager 2 IP address)

Called Phone: 6109565

Time stamp: 11:53:36


Why Cisco SME is misbehaving ?


Please let me know if moer information required.


Thanks in advance !!!


Best Regards,

Suresh

Correct Answer by Ayodeji Okanlawon about 3 years 6 months ago

Suresh,


I will advise to schedule a time to test the secondary alcatel server by removing the primary sip trunk from the RG. I am sure that this server is not working or the connection to it is not working.


The service parameter is used for H323 gateway redundancy purposes especially with E1 circuits. E,g. When you have a two H323 gateways configured with E1s. When the E1 on the primary gateway  goes down, the gateway will send a user busy flag to CUCM. When CUCM receives this, what it will do next depends on what this parameter is set to. If it is set to "TRUE", then cucm will drop the call, if it is set to "FALSE" then CUCM will send the call to the secondary gateway in the RG.


This behaviour is similar with CUBEs and SIP trunks. When CUCM receives a 486 user busy or 401 unallacated or un assigned number, it will use this parameter to determine what it does next.


This parameter was built into the SIP stack in CUCM to allow for redeundancy purposes as well. e.g. Supposing you have multiple cluster in your environment and in your RG, you have configured two sip trunks, one to the first alcatel cluster and the second to another alcatel cluster. When calls to the first cluster receives busy, this will allow SME to send the call to another cluster.


I believe it is safe to say that you can leave this parameter set to True, fi all you have is a SIP trunk. I dont see a need for SME to try and send a call that was either rejected by the end user or busy to another place


Please rate all useful posts


"The essence of christianity is not the enthronement but the obliteration of self --William Barclay"

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (6 ratings)
Loading.
Leo Salcie Tejeda Sat, 02/01/2014 - 15:52
User Badges:
  • Gold, 750 points or more

Did you mean CME( call manager express) or SME( Storage Media Encription)?

Suresh Hudda Sat, 02/01/2014 - 20:52
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Hi Leosalcie,


It is Cisco Unified CM Session Management Edition.


Regards,Suresh

Manish Prasad Mon, 02/03/2014 - 02:30
User Badges:

Hi Suresh,


Can you attach only "debug ccsip message" logs , it will be more easy to read it.


Thanks

Manish

Suresh Hudda Mon, 02/03/2014 - 05:04
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Hi Manish,


It is not installed on router. It is SME which installed on MCS 78XX or UCS and there we cant execute such command.


Regards, Suresh

Suresh Hudda Mon, 02/03/2014 - 05:06
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Dear Experts,


Any suggestion please ?


Regards, Suresh

Suresh Hudda Sat, 02/08/2014 - 10:06
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Finally I did it...I mean it has resolved now by changing the service parameter " stop routing on user busy flag" to True !!!

Ayodeji Okanlawon Sat, 02/08/2014 - 15:57
User Badges:
  • Super Bronze, 10000 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Suresh,


I just saw your post today and I got interested! I wanted to know why SME didnt send the 486 to Alcatel. My findings opened some things you may need to check...Here are my thoughts..


++++Alcatel sent a 486 Busy to SME+++


11:53:37.094 |//SIP/SIPTcp/wait_SdlReadRsp: Incoming SIP TCP message from 10.205.228.9 on port 5060 index 2507 with 473 bytes:

SIP/2.0 486 Busy Here

Allow: INVITE, ACK, CANCEL, BYE, PRACK, NOTIFY, SUBSCRIBE, OPTIONS, UPDATE

User-Agent: OmniPCX Enterprise R10.1.1 j2.603.25.a

To: ;tag=24b9474f6e8f0cab4601659e60dd9c1a

From: "Aishwarya G" ;tag=f4e2ae3b-5545-4b09-bba6-7d25c800704e-24066665

Call-ID:

[email protected]


+++Next SME sent an ACK+++++


11:53:37.094 |//SIP/SIPTcp/wait_SdlSPISignal: Outgoing SIP TCP message to 10.205.228.9 on port 5060 index 2507

ACK sip:[email protected]:5060 SIP/2.0

Via: SIP/2.0/TCP 10.205.25.10:5060;branch=z9hG4bK6d04474b92a

From: "Aishwarya G" ;tag=f4e2ae3b-5545-4b09-bba6-7d25c800704e-


Now what happens next is the reason SME doesnt relay the 486 to Alcatel..


+++SME sends a setup request to another device in the RL++++


11:53:37.095 |RouteListCdrc::null0_CcSetupReq - Selecting a device.|1,100,56,1.61627^10.205.228.9^*


+++SME found RG and the Device in the RG++++


1:53:37.095 |SMDMSharedData::findLocalDevice - Name=GGN_SME_To_BLR_ALCATEL

11:53:37.095 |RouteListCdrc::executeRouteAction: EXTEND_CALL_TO_CURRENT_MEMBER -- Success

11:53:37.095 |ViprUtils:isViprAllowed  Device =GGN_SME_To_CHE_ALCATEL_8  UseIMEForOutboundCall=true


++++Next SME sends a TCP connection request to the gateway (10.205.192.178) on Port 5060+++++


11:53:37.100 |//SIP/Stack/Transport/0xe456120/Sending Invite to the transport layer|1,100,56,1.61627^10.205.228.9^*

11:53:37.100 |//SIP/Stack/Transport/0xe456120/msg=0xb39588e8, addr=10.205.192.178, port=5060

11:53:37.100 |//SIP/Stack/Transport/0x0/gConnTab=0xb7a5cf90, addr=10.205.192.178, port=5060, connid=2515, transport=TCP|1,100,56,1.61627^10.205.228.9^*

11:53:37.100 |//SIP/Stack/Transport/0x0/Moving connection=0xb7a5e808, connid=2515state to pending|1,100,56,1.61627^10.205.228.9^

11:53:40.982 |//SIP/SIPUdp/wait_UdpDataInd: Incoming SIP UDP message size 468 from 10.205.39.8:[5060]:

CANCEL sip:[email protected];user=phone SIP/2.0

Supported: timer,path,100rel

User-Agent: OmniPCX Enterprise R10.1.1 j2.603.25.a

Call-ID:

[email protected]

11:53:37.100 |//SIP/Stack/Transport/0xe456120/Sending Invite to the transport layer|1,100,56,1.61627^10.205.228.9^*
11:53:37.100 |//SIP/Stack/Transport/0xe456120/msg=0xb39588e8, addr=10.205.192.178, port=5060

11:53:37.100 |//SIP/Stack/Transport/0x0/Posting TCP conn create request for addr=10.205.192.178, port=5060

++++++However there is a problem with the TCP connection, SME moves the connection to a pending state+++++++


11:53:37.100 |//SIP/Stack/Transport/0x0/gConnTab=0xb7a5cf90, addr=10.205.192.178, port=5060, connid=2515, transport=TCP|1,100,56,1.61627^10.205.228.9^*
11:53:37.100 |//SIP/Stack/Transport/0x0/Moving connection=0xb7a5e808, connid=2515state to pending|1,100,56,1.61627^10.205.228.9^


++++Next 3secs after Alcatel sent a 486 busy, it sends a CANCEL+++++


11:53:40.982 |//SIP/SIPUdp/wait_UdpDataInd: Incoming SIP UDP message size 468 from 10.205.39.8:[5060]:

CANCEL sip:[email protected];user=phone SIP/2.0

Supported: timer,path,100rel

User-Agent: OmniPCX Enterprise R10.1.1 j2.603.25.a

Call-ID:

[email protected]


Summary:


Cisco CUCM, CUBE by default do not drop a call when it recieves a 486 busy, 404 not found it will try another destination if one exists. This is what happened here. SME found another device to send the call to and it attempted to send it there, however the TCP connection to that device failed.

You should investigate why your second device in that RG is not accepting calls.


Your work around obviously worked because its telling SME not to route call to the next device when it gets a busy, but its hiding your real problem.




Please rate all useful posts


"The essence of christianity is not the enthronement but the obliteration of self --William Barclay"

Suresh Hudda Sat, 02/08/2014 - 22:19
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Hi Aokanlawon,


Thanks a lot for the nice observation, You deserve +5 rating.. Actually there are two trunks (GGN_SME_To_BLR_ALCATEL_9 && GGN_SME_To_BLR_ALCATEL) in RG which pointing to Alcatel, one pointing to Alcatel primary Call manager and another trunk is pointing to standby call manager.


Now I need to confirm with Alcatel engineer that, whether standby callmanager can process calls when primary call manager is active.


But in my scenerio if end user's phone is busy or not available then SME should not try to reach another member in RG.


I mean what would be the impact if I kept service parameter "stop routing on user busy flag" to true ? Bydefault this setting is "true" only, dont know why it was set to false.


Regards, Suresh




Correct Answer
Ayodeji Okanlawon Sun, 02/09/2014 - 01:29
User Badges:
  • Super Bronze, 10000 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Suresh,


I will advise to schedule a time to test the secondary alcatel server by removing the primary sip trunk from the RG. I am sure that this server is not working or the connection to it is not working.


The service parameter is used for H323 gateway redundancy purposes especially with E1 circuits. E,g. When you have a two H323 gateways configured with E1s. When the E1 on the primary gateway  goes down, the gateway will send a user busy flag to CUCM. When CUCM receives this, what it will do next depends on what this parameter is set to. If it is set to "TRUE", then cucm will drop the call, if it is set to "FALSE" then CUCM will send the call to the secondary gateway in the RG.


This behaviour is similar with CUBEs and SIP trunks. When CUCM receives a 486 user busy or 401 unallacated or un assigned number, it will use this parameter to determine what it does next.


This parameter was built into the SIP stack in CUCM to allow for redeundancy purposes as well. e.g. Supposing you have multiple cluster in your environment and in your RG, you have configured two sip trunks, one to the first alcatel cluster and the second to another alcatel cluster. When calls to the first cluster receives busy, this will allow SME to send the call to another cluster.


I believe it is safe to say that you can leave this parameter set to True, fi all you have is a SIP trunk. I dont see a need for SME to try and send a call that was either rejected by the end user or busy to another place


Please rate all useful posts


"The essence of christianity is not the enthronement but the obliteration of self --William Barclay"

Suresh Hudda Sun, 02/09/2014 - 02:39
User Badges:
  • Gold, 750 points or more
  • Cisco Designated VIP,

    2017 IP Telephony

Thanks Aokanlawon once again, sure I will test calling removing primary trunk and will se the behaviour. In SME we are using only SIP trunks so we will keep this parameter true.


Regards, Suresh

Actions

This Discussion