Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. And see here for current known issues.

H323/MGCP - FXS (Fax/Modem)

Have two voice gateways (3845) terminating PRIs in an NFAS configuration. So the router's config is H323 based. Have 6 VG224s deployed in H323 as well, because a majority of the devices involved are fax and modems. Since forced authrorization code is a requirement for dialing long distance calls (even from fax machines and modems), running into an issue with H323 on the VG224. Analog gateways cannot play tones required for forced auth codes and hence FAC/CMC are not supported with H323.

So I changed the configs from H323 to MGCP on the VG224s (my PSTN gateways are still H323 due to NFAS). After this change, FAC codes seem to work on analog devices properly. I had enabled Cisco fax relay initially on the PSTN gateway (h323) and MGCP enabled VG224. This didnt seem to help much, so i changed the configs to use T38 fax relay. Fax seem to start working immediately after i changed to T38. But this behaviour seems to be very inconsistent. Sometimes it work, sometimes it doesnt.

Configs on H323 PSTN gateway..


dial-peer voice 2 voip

description Dialpeer to Match Incoming Voip leg

incoming called-number .

dtmf-relay h245-alphanumeric

codec g711ulaw

fax-relay ecm disable

fax protocol t38 ls-redundancy 0 hs-redundancy 0 fallback cisco

ip qos dscp cs3 signaling

no vad


Configuration of MGCP VG224.


ccm-manager mgcp

no ccm-manager fax protocol cisco



mgcp call-agent <CM-ip> 2427 service-type mgcp version 0.1

mgcp dtmf-relay voip codec all mode out-of-band

mgcp rtp unreachable timeout 1000 action notify

mgcp ip qos dscp cs3 signaling

mgcp package-capability rtp-package

no mgcp package-capability res-package

mgcp package-capability sst-package

no mgcp package-capability fxr-package

no mgcp timer receive-rtcp

mgcp sdp simple

<b>no mgcp fax t38 ecm

no mgcp fax t38 inhibit (default)</b>

mgcp rtp payload-type g726r16 static

VG224 is running 12.4.12 (LD).

The H323 gateway is a 3845 running 12.3.14T7. The reason this gateway is running this version due to a bug with NFAS configuration in the 12.4 code (all of them are affected, so we had to downgrade to 12.3 until a fix for it arrives).

Any suggestions or improvements in the config above to make this work ?


Hall of Fame Super Gold

Re: H323/MGCP - FXS (Fax/Modem)

Hi Shanky,

at first I didn't realized it was you posting this problem. I cannot really help but I have a couple of questions.

Why are you running ecm disabled? There are bugs in there too?

What if you push cisco really hard on providing a fix for for nfas bug in 12.4? if you could move ahead instead than backward in release that would be a plus.

Have you ever heard of cisco's CA (critical account) status? Basically is when you have one or more P1/P2 cases that are impacting your business, and the whole business relationship is about to be compromised (ie, cisco risks loosing the account forever) because something in the product is not working right.

Just food for thought, hope can help.

Re: H323/MGCP - FXS (Fax/Modem)


Faxing seem to be working now consistently. The issue i was having seem to be an isolated issue with a few fax machines. I had ecm disabled because this is a lot easier to do than individually going into each fax machine and disabling ECM. Technically there is no reason to disable ECM, unless your network has packet loss > 2%.

The NFAS bug was a hard one to troubleshoot, took a week to figure out downgrading to 12.3 will fix the problem. We are at the last release of 12.3 (12.3.14.T7) so we are currently not losing any features. But that doesnt of course mean we are free of bugs in the 12.3 code.

Past 6 months have been so grueling with this project that I havent really thought about going back to Cisco and press them on this issue. A lot of escalation happened with the NFAS issue and thats the reason why Cisco immediately suggested going to 12.3. Lucky for us, no other issues have started to pop up. The case is still open with Cisco and as soon as a fix is available, the engineer has promised to let me know.