H323 failover between two routers

Unanswered Question
Apr 16th, 2012
User Badges:


We have below call flow configured:-

IP Phone (IP phone ext. number 8061, called number 09930907777) -----> CUCM ------> Router-2 (h323 gateway, its PRI is down. Call routed to Router-1 by dial peer)-------> Router-1(h323 gateway, PRI is up call is routed to service provider) --------> Mobile phone (Phone number 9930907777 rings for around 10 seconds and gets disconnected)

Description:- I have CUCM 8.X and 2 voice gateway router configured on h323 protocol. Both the router has PRI connectivity. CUCM is configured so that it routes call first to Router -2, if router-2 PRI is down the call is routed to router-1 by the dial-peer configured to router-2. The call is sent to service provider using PRI terminated on router-1

Problem:- In above scenarion when router-2 PRI is down, and the IP Phone (ext number 8061) calls the mobile number (09930907777) the mobile phone rings for only 10 seconds and call gets disconnected. If mobile user answers the call within 10 seconds then also call gets disconnected.

Attached is show run of both the routers. Also find attached file with output of:-

debug isdn q931

debug voice dial-peer inout

debug voice ccapi inout

debug cch323 all

Please suggest what is reason for this problem and its resolution.

Thanks in advance.



  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
marmugam Mon, 04/23/2012 - 20:48
User Badges:
  • Cisco Employee,


Sounds like Media Exchange timeout in CCM Service parameter (Mx timeout), Ideally it is 12 seconds. Just to isolate ,whether that is the issue, can you increase the timeout and test the call ? If it works let me know,  I am susepcting another thing,

If it doesn't get debug isdn q931 from the router and Detailed CCM manager logs will help.



marmugam Mon, 04/23/2012 - 21:06
User Badges:
  • Cisco Employee,

++ OK, Looked into the Router log attached, Cant be MX timeout as Alerting is recevied @ 12:33:40 witn inband information(PI=8), so even if CCM started  MX timer, we should see a disconnect only after 12 secs.

However, I am not sure you did went onhook in this case. Detailed CCM logs will give more details.

Apr 13 21:33:40.367: ISDN Se1/0:15 Q931: RX <- ALERTING pd = 8  callref = 0xEB49

Progress Ind i = 0x8288 - In-band info or appropriate now available .

Apr 13 21:33:47.683: ISDN Se1/0:15 Q931: TX -> DISCONNECT pd = 8  callref = 0x6B49

Cause i = 0x80AF - Resource unavailable, unspecified

++ Also when it is not failing over, do you still see the issue when the cell phone user didnt go offhook in 10 secs ? Or is it consistently happenning in the failover issue ?

acampbell Tue, 04/24/2012 - 01:37
User Badges:
  • Green, 3000 points or more


I think this will help you



no dial-peer outbound status-check pots



This line of config will sent back no circuit/chanel available to CUCM allowing it to step on through a route group if dial peers are down

See link


Usage Guidelines

Use this command to disallow, during call setup, outbound POTS dial peers (except those for e-phones) whose endpoints (voice ports or trunk groups) are down.

When the dial-peer outbound status-check pots command is configured, if the voice-port configured under an outbound POTS dial-peer is down, that dial-peer is excluded while matching the corresponding destination-pattern. Therefore, if there are no other matching outbound POTS dial-peers for the specified destination-pattern, the gateway will disconnect the call with a cause code of 1 (Unallocated/unassigned number), which is mapped to the "404 Not Found" SIP response by default. When the no form of this command is configured, the outbound POTS dial-peer is matched even if the voice-port configured under is down and the gateway disconnects the call with a cause code of 34 (No circuit/channel available), which is mapped to the "503 Service Unavailable" SIP response by default.



andersson1970 Tue, 04/24/2012 - 03:01
User Badges:

Hi Acampbell,

I configured command "no dial-peer outbound status-check pots"  on outbound dial-peer, then I shut the PRI on the router to simulate failover. After this when I dialled the PSTN number from IP phone I got cucm announcement as "Your call can not be completed ...."

Thus with this commnad unable to achieve failover to second h323 gateway configured in route group.

Hi Marmugam,

In case of normal scenario call works properly and does not disconnect after 12 sec. I am facing the problem only while failover, that call get disconnected after 12 Sec.

Please suugest to resolve the problem.



marmugam Tue, 04/24/2012 - 08:33
User Badges:
  • Cisco Employee,


This is the call flow per my understanding, correct me if I m wrong:-

IP phone--CCM---h225----Router 2-----PRI down




                                       Router 1----PRI------Telco----Mobile Phone.

In this scenario, CCM doesn't know PRI is down, so it selects the Device sends the call out.  Router 1 will not even attempt to send the call to PRI, as it is down, so CCM receives h225 disconnect, CCM upon receiving the message looks into the cause code and decides whether to hunt the remaining devices in the list or not, in this case,  it is hunting down the list, and selcts ROuter2 and sends the call out.

So I don't think any dial-peer config here is necessary.  (debug isdn q931/debug h225 asn1 and debug h245asn1) from router 2 and detailed CCM logs will let us know what is happening. Please attach them, I will defintely take a look.


oliverpowell Thu, 01/22/2015 - 07:26
User Badges:

Hi Anderson,


Did you ever resolve this issue of h323 failover? We are having a similar problem with one of our depoyments. We've migrated from MGCP to H323 gateways on CUCM.




vincent.morton Sun, 05/01/2016 - 10:33
User Badges:


I realise this is an old post and you've probably already found the answer but in case it's useful to anyone else having the same problem, if outbound calls fail to fail over to another E1 on H323 gateways when there is a fault with the first ISDN in the outbound list, there are two service parameters to check on CUCM:

"Stop Routing on Out of Bandwidth Flag"

Default: True

"Stop Routing on User Busy Flag"

Default: True

Setting these parameters to false will allow failover to a second route group member if the 1st cannot complete the call.


This Discussion

Related Content