cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1193
Views
13
Helpful
9
Replies

Call routing to PSTN based on call origin

vigleik
Level 1
Level 1

I have a CM 4.1 cluster with several sites.

In Site_A and Site_B there is a Callmanager and a gateway with ISDN PRI to the PSTN.

I have configured two Callmanager groups, one with CM_A as primary and CM_B as secondary.

The other CM group has CM_B as primary and CM_A as secondary.

Phones register to the closest Callmanager based on the CM group of their device pool.

If one callmanager goes down, phones fail over to the other.

There is a 10 Mbps link between the sites, so bandwidth limitation is no issue.

Now the question:

I want calls to the PSTN to be routed through the nearest gateway.

That is, phones in Site_A should use GW_Site_A as first choice. If all lines are busy, GW_Site_A should be used.

In Site B the other way around.

How can that be done ? Do I need to configure locations ?

1 Accepted Solution

Accepted Solutions

Chris Deren
Hall of Fame
Hall of Fame

Here is what you need to do:

create location specific paritions:

brA-emergency-pt

brA-local-pt

brA-ld-pt

brA-int-pt

brB-emergency-pt

brB-local-pt

brB-ld-pt

brB-int-pt

create CSS to be assigned to phones based on calling privilages, i.e.

brA-unrestricted-css

- internal-pt (if you have one)

- brA-emergency-pt

- brA-local-pt

- brA-ld-pt

- brA-int-pt

Create route groups

brA-RG assign branch A GW it it

brB-RG assign branch B GW to it

Create route lists

brA-RL

-brA-RG

-brB-RG

Create route pattern, i.e. for long distance calling:

9.1[2-9]XX[2-9]XXXXXX in partition brA-ld-pt

point this RP to brA-RL

HTH, please rate all posts!

Chris

View solution in original post

9 Replies 9

Chris Deren
Hall of Fame
Hall of Fame

Here is what you need to do:

create location specific paritions:

brA-emergency-pt

brA-local-pt

brA-ld-pt

brA-int-pt

brB-emergency-pt

brB-local-pt

brB-ld-pt

brB-int-pt

create CSS to be assigned to phones based on calling privilages, i.e.

brA-unrestricted-css

- internal-pt (if you have one)

- brA-emergency-pt

- brA-local-pt

- brA-ld-pt

- brA-int-pt

Create route groups

brA-RG assign branch A GW it it

brB-RG assign branch B GW to it

Create route lists

brA-RL

-brA-RG

-brB-RG

Create route pattern, i.e. for long distance calling:

9.1[2-9]XX[2-9]XXXXXX in partition brA-ld-pt

point this RP to brA-RL

HTH, please rate all posts!

Chris

Thanks a lot, exactly what I needed!

I'm new to voice gateways, so I have a couple of more questions.

The gw in site B is a new 2811 voice bundle with an ISDN Pri E1 interface.

The existing gw in site A is a VG200 running H323.

I guess it's OK to use MGCP for the new one and H.323 for the old one.

I have not yet connected the ISDN PRI in site B because it is in production in an old PBX.

The mgcp gateway is registered, but the mgcp endpoints has state "not registered" in CM.

- Is the "unregistered" state normal for the mgcp endpoint as long as the physical port is not connected ?

We will do a short off hours test to try inbound and outbound calls through the mgcp gateway.

- How can I trace a call in CallManager to see what happens when numbers are dialed ?

I want to look what happens also before connecting the line.

The Dialed number analyzer just shows me what route list will be used, nothing more.

What debug commands are useful in the router ?

- When a site A phone calls PSTN through the site B GW, will the ISDN PRI provider accept a caller id not belonging to that ISDN line ?

- When running H.323 for VG200, Will CM know if the ISDN PRI goes down, so all external calls can go through site B based on RL_Sovik?

Vigleik

There is no issues with mixing MGCP and H.323 GWs.

The status of the GW is OK, if you can please post your config and "sh ccm-manager".

As to tracing the calls, in CCM you can use the CCM trace logs, but unless you have experiece reading these it might be challanging.

The best way to trace the call on ISDN PRIs is to use "debug isdn q931", you will see exactly what number is being dialed, disconnect reason codes, which channel, etc.

Caller ID shoud be fine as longs as your service provider is providing it, and you are not blocking it on Route Patterns.

CCM will know that VG200 is down and will select the next Route Group within the Route List that was invoked.

HTH,

Chris

Here's the config and the output.

I think I can delete the two voip dial peers but must keep one for pots , is that right ?

Can I also delete the num-exp commands ?

I think I need another DSP module to be able to use all 30 B-channels simultaneously.

Correct ?

Vigleik

From the config:

- -

hostname c2811

card type e1 0 0

!

resource policy

!

network-clock-participate wic 0

!

ip cef

isdn switch-type primary-net5

!

voice-card 0

no dspfarm

!

voice service voip

!

voice class codec 1

codec preference 1 g711alaw

codec preference 2 g711ulaw

codec preference 3 g729r8

codec preference 5 gsmfr

!

controller E1 0/0/0

pri-group timeslots 1-16 service mgcp

description 2Mbit PRI til PSTN

!

interface FastEthernet0/0

ip address 172.17.32.2 255.255.252.0

!

interface Serial0/0/0:15

no ip address

isdn switch-type primary-net5

isdn incoming-voice voice

no cdp enable

!

ip classless

ip route 0.0.0.0 0.0.0.0 172.17.32.1

!

control-plane

!

voice-port 0/0/0:15

!

ccm-manager redundant-host 172.23.2.207

ccm-manager mgcp

!

mgcp

mgcp call-agent 172.23.1.208 service-type mgcp version 0.1

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

mgcp rtp unreachable timeout 1000 action notify

mgcp modem passthrough voip mode nse

mgcp package-capability rtp-package

mgcp package-capability sst-package

no mgcp timer receive-rtcp

mgcp sdp simple

no mgcp explicit hookstate

mgcp bind media source-interface FastEthernet0/0

!

mgcp profile default

!

dial-peer voice 100 pots

description Calls to PSTN

destination-pattern 0..

progress_ind setup enable 3

progress_ind alert enable 8

progress_ind progress enable 8

direct-inward-dial

!

dial-peer voice 1 voip

preference 1

destination-pattern 83...

progress_ind setup enable 3

progress_ind progress enable 8

voice-class codec 1

session target ipv4:172.23.1.208

incoming called-number .

fax nsf 000000

no vad

!

dial-peer voice 2 voip

preference 2

destination-pattern 83...

progress_ind setup enable 3

progress_ind progress enable 8

voice-class codec 1

session target ipv4:172.23.2.207

incoming called-number .

fax nsf 000000

no vad

!

num-exp 77783... 83...

num-exp 77771... 71...

! The 71... series will be added later, on this same PRI

!

scheduler allocate 20000 1000

c2811# show ccm-manager

MGCP Domain Name: c2811

Priority Status Host

Primary Backup Ready 172.23.1.208

First Backup Registered 172.23.2.207

Second Backup None

Current active Call Manager: 172.23.2.207

Backhaul/Redundant link port: 2428

Failover Interval: 30 seconds

Keepalive Interval: 15 seconds

Last keepalive sent: 20:40:57 UTC Oct 18 2006 (elapsed time: 00:00:03)

Last MGCP traffic time: 20:40:57 UTC Oct 18 2006 (elapsed time: 00:00:03)

Last failover time: 20:40:27 UTC Oct 18 2006 from (172.23.1.208)

Last switchback time: 13:55:14 UTC Oct 17 2006 from (172.23.2.207)

Switchback mode: Graceful

MGCP Fallback mode: Not Selected

Last MGCP Fallback start time: None

Last MGCP Fallback end time: None

MGCP Download Tones: Disabled

Configuration Error History:

FAX mode: cisco

c2811#sh mgcp endp

Interface E1 0/0/0

ENDPOINT-NAME V-PORT SIG-TYPE ADMIN

S0/SU0/ds1-0/1@c2811 0/0/0:15 none up

S0/SU0/ds1-0/2@c2811 0/0/0:15 none up

...and so on...up to 15

S0/SU0/ds1-0/15@c2811 0/0/0:15 none up

Your MGCP config looks good. One thing to

Which PVDM module do you have in your router? PVDM2-16? If that's the case then you'll need more DSPs.

Since you only enabled 16 channels, make sure that the service provider is sending the calls in ascending order, so the first call comes in on channel 1 not 30.

If you are not doing H.323 or SRST on this GW then you don't need any of the dial-peers including the pots ones, with PRIs mgcp pots dial-peers are not needed.

The num-exp will also not be used, unless you use H.323 or SRST on this GW.

Chris

Yes, it's a PVDM2-16.

Another one is ordered.

Tonight we connected the ISDN PRI to the VWIC2-1MFT-T1/E1 to test.

In Callmanager, the endpoints then show as registered.

In the router, I got these two error messages continuously from "debug isdn q931":

ISDN Se0/0/0:15 **ERROR**: L2IF_SendPkt: idb is NULL

ISDN Se0/0/0:15 **ERROR**: process_rxdata:L2IF_SendPkt Failed

No calls can be made.

What's wrong ?

The config looks like this:

----------------

card type e1 0 0

!

resource policy

!

network-clock-participate wic 0

ip subnet-zero

!

ip cef

!

isdn switch-type primary-net5

!

voice-card 0

no dspfarm

!

voice service voip

!

voice class codec 1

codec preference 1 g711alaw

codec preference 2 g711ulaw

codec preference 3 g729r8

codec preference 5 gsmfr

!

controller E1 0/0/0

pri-group timeslots 1-31 service mgcp

description 2Mbit PRI til PSTN

!

!

!

interface Serial0/0/0:15

no ip address

isdn switch-type primary-net5

isdn incoming-voice voice

no cdp enable

!

control-plane

!

voice-port 0/0/0:15

!

ccm-manager redundant-host x.x.x.x

ccm-manager mgcp

!

mgcp

mgcp call-agent y.y.y.y service-type mgcp version 0.1

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

mgcp rtp unreachable timeout 1000 action notify

mgcp modem passthrough voip mode nse

mgcp package-capability rtp-package

mgcp package-capability sst-package

no mgcp timer receive-rtcp

mgcp sdp simple

no mgcp explicit hookstate

mgcp bind media source-interface FastEthernet0/0

!

mgcp profile default

!

You are missing:

"isdn bind-l3 ccm-manager" under interface Serial0/0/0:15

Chris

Thanks, I added the command and we're making progress. Today's test leaves us with two problems:

Problem 1 - Incoming calls work ok if they are answered within 5 seconds. If not, the call is dropped. Perhaps some error in the CM gateway configuration, which is shown in the attachments ? Do I need to ask the ISDN provider?

Problem 2 - the leading zero is not stripped off outgoing calls. Changed "number of digits to strip" from 0 to 1. Is that the way to do it? It didn't help, but that may be because the gw was not reset properly, as the debug shows. Bringing down and up the serial interface may not be enough ?

The attachments show a debug isdn q931 of incoming and outgoing calls. The jpg files show the endpoint config from CCM. Was wondering about "called numbering plan" for outgoing, but it's set to "Callmanager" for a working H323 gateway.

Vigleik

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: