cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
4258
Views
0
Helpful
14
Replies

MGCP gateway problem - channels stucked?

Sinisa Hreljac
Level 1
Level 1

Hi!

I have a problem with MGCP gateways - from time to time they cannot receive or send any calls. Then I have to shutdown/no shut E1 controller and then for some time it works OK.

Investigating this problem using RTMT PRIChnnelsActive performance monitor and looking at CLI on MGCP gateway with "show mgcp connections" I see problem in wrong count number of active channels (like on attached picture).  On RTMT (in this case) it shows 18 active channels while on router it shows only 5. What is happening here is that after E1 reset , this numbers are OK on RTMT and in CLI, but after some time (few days, weeks) it starting to mismatch, showing always bigger number in RTMT. This number in RTMT grows until it reaches all available channels and than no more calls can be done.

While active calls in RTMT growing, number of calls that can be acctualy made (that can be shown in CLI with "sh mgcp connections") gets down, so sometimes can be maximum of 2 active calls, while RTMT shows 30. So in my conclusion, there is some problem with stucking channels - in case of attached picture, there is 18 - 5 = 13 channels that are stucked. They cannot be seen with "sh mgcp connections" on router, but are shown on RTMT.

In this case, max number of active channels can be 30 - 13 = 17.

This is happening on all MGCP gateways in our VoIP network - more than 10 of them; some of them are in different locations (towns), some of them are with IOS 12.4, some with 15.0...

We are using CUCM 7.1.5.

Becouse of this problem sometimes calls to or from our company not working until we manually reset E1 controller.

So question is, how to troubleshoot this case? Can this be problem on telco provider side (E1 side) or internal in our system (routers to CUCM)?

Can I somehow view this stucked channels on MGCP gateway, not only via RTMT and using maths - number of channels that are shown in RTMT minus number of channels shown via "sh mgcp connections command" on router?

Thanks and best regards,

SH.

14 Replies 14

Ayodeji Okanlawon
VIP Alumni
VIP Alumni

Hi,

What version of CUCM are you running. This behaviour looks similar to this BUG

CSCta77785 - MGCP port status not communciated to CUCM well

Symptom:
MGCP ports may be marked with an incorrect status (idle/busy) on the CUCM. This causes calls to be presented to ports it can't traverse, resulting in a fast-busy or causes calls to be rejected that could be put thru.
Conditions:
MGCP CAS/PRI trunks with MLPP.
Workaround:
shut/noshut on the controller allows the channels to resync, but obviously interrupts service.Is this bug description helpful?

(0)DetailsFirst Found in: (1)
8.6(0.98000.9)
More
LessStatus:Fixed (Resolved)Last Modified:Jan 13,2012Known Affected Versions: MoreLess
Fixed in: (18)
9.0(0.96000.1),9.0(0.95070.39),9.0(0.95070.38)
9.0(0.95010.1),8.6(1.98000.43),8.6(1.96000.16)
8.6(1.95050.1),8.6(1.95020.80),8.6(1.95020.1)8.6(1.10001.1),8.6(1.10000.43),8.6(1.10000.1)8.6(0.99981.2),8.6(0.98000.46),8.6(0.95180.9)8.5(1.12021.1),8.0(3.23032.1),8.0(2.40006.1)

Please rate all useful posts

"There is a wideness in God's mercy Like the wideness of the sea.There's a kindness in His justice Which is more than liberty"

Please rate all useful posts

Sinisa Hreljac
Level 1
Level 1

Hi,

Cucm version is 7.1.5.34094-1

Is this version also affected with this bug?

Sent from Cisco Technical Support iPhone App

The bug history does not say it affects your version. I will open a TAC case with cisco to see if its the same bug.

Please rate all useful posts

"There is a wideness in God's mercy Like the wideness of the sea.There's a kindness in His justice Which is more than liberty"

Please rate all useful posts

Hi aokanlawon,

is there any news about this?

Thanks!

Hi,

I advised you to Open a TAC case with Cisco. Did you open one?

Please rate all useful posts

"'Nature is too thin a screen, the glory of the omnipresent God bursts through it everywhere"-Ralph Waldo Emerson

Please rate all useful posts

Hi,

I obviously missunderstand you; you said "I will open a TAC case with cisco to see if its the same bug." so I thought that you will be check with TAC for this case - which was strange to me in first place (I thought why would you open TAC case for problem that is not yours).

However, currently we have no Cisco support (it is in renewal process) so in meanwhile I trying to find workaround for this issue.

I tried to setup nightly controller reset on routers, but it does not work. In logs I can see just my syslog comment (so job is scheduled and working OK) but no controller reset.

Here is config:

event manager applet reset-E1

event timer cron name reset-E1-schedule cron-entry "10 01 * * *"

action 1.0 cli command "enable"

action 1.1 cli command "controller e1 1/0"

action 1.2 cli command "shutdown"

action 1.3 cli command "no shutdown"

action 1.4 syslog msg "E1 controller reseted"

In sh log I see only this:

833987: Jul 25 01:10:00.777 UTC: %HA_EM-6-LOG: reset-E1: E1 controller reseted

There is no controller shutdown/no shutdown or Serial interfaces UP/DOWN events.

Is there a way to schedule nightly controller reset?

Best regards,

S.

I dont think there is any way to automate the process. Have you considered upgrading your IOS. Thats another option. Since you dont have a contract I will suggest you upgrade your IOS. Try that and see if it helps

Please rate all useful posts

"'Nature is too thin a screen, the glory of the omnipresent God bursts through it everywhere"-Ralph Waldo Emerson

Please rate all useful posts

Just on a slightly different note.

is there anyway to reset only a single ISDN channel instead of the whole E1.

like we had the first channel on the first gateway out of service (see attachement) and there were active calls on the rest of the trunk.

so just to bounce one single channel do we really have to do the whole trunk!!!

Also at that time, the show isdn service or the show voice port summary doesnt show any issues with the channel. Only can be found in the RTMT

thanks

Hi abbas_wali,

maybe stupid question, but I cannot find nowhere in RTMT screen like you attached. Where can I find screen like this to see which channels are used and which are out?

Thanks!

you can get that from

CallManager>Device>Device Search>Open Device Search>Gateway Devices

Then look for the Gateways which depends are what you have implimented in your cluster. we have MGCP Trunks here.

It will show you the cards registered with the CM adn if you right click on them you will see port status.

hope that helps.

Ta

Abbas

Hi SH,

Have got it solve this problem ? We are also encountering the SAME issue. But we are using CUCM 8.5 and MGCP VG IOS is 12.4. This is started to happen few month before, had to reset the MGCP VG to tally the active call count on the RTMT and VG.

tks

ragu

Hi ragu,

sorry for  very long delay in answer.


We did not solve problem. Still in communication with TAC and they still haven't any clue why is this happening.

Did you found solution?

If not, maybe we can contact each other to exchange details about CUCM configuration and scenarios when this problem happens.

SH.

Hi,

I have the similar problem.

regds,

aman

You can contact me on sinisa.hreljac@s-itsolutions.hr for experiance exchange on this issue.

Tnx.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: