trying to configure survivability.tcl but no luck

[Luis Yrigoyen]

Hi,

I'm trying to configure the survivability.tcl service but I can't seem to get it to work.  I've followed the instructions on the CVP config/admin guide but no luck.  Attached is the config from a 2811 I'm using on the LAB.

My setup:

ICM 8.5 server running Rogger, Logger, PG, CTIOS, HDS (x.x.x.82)

CVP 8.0 server running CVP call server, vxml server, media server (x.x.x.83)

UCM 8.6 call manager server (x.x.x.81)

2811 running (C2800NM-IPVOICEK9-M) 12.4(24)T5 (x.x.x.254)

     4 FXO ports

I would appreciated if someone could look at the config and tell me what's wrong or point me to documentation.

The config is in the attached file.

thanks

[geoff · Accepted solution]

You apply the service to the pots dial peer - a bit like you have:

dial-peer voice 3 pots

service survive

incoming called-number 4930

direct-inward-dial

I've not seen anyone write "service survive out-bound" - not sure if you can add a parameter "out-bound" like that.

So when you call 4930, the call is the under control of the survivability service. It applies a timer and if CVP cannot be contacted, survivability can forward the call to a phone that has failed over to call-manager-fallback - we expect that the reason CVP cannot be contacted is that the WAN is down, so the phones have failed over to SRST. You can set up the service like this:

service cvp-survivability flash:survivability.tcl

  paramspace english language en

  paramspace english index 0

  param open-hours-time0 0123456:0700-2359

  paramspace english location flash

  param open-hours-agent0 23507

  param open-hours-agent1 23504

  param open-hours-agent2 23503

It will go down the list of agents, like a hunt group, and as expected from agent0 to agent2. You get a female voice saying "please wait while I transfer you" followed by music on hold while the call cycles through the agents.

Looking at your gateway config, I see a problem.

dial-peer voice 99 voip

  description For Incoming Leg (Type 10 label and Correlation ID)

  service bootstrap

  incoming called-number 9999999999

  dtmf-relay rtp-nte h245-signal h245-alphanumeric

  codec g711ulaw

  no vad

You are not allowing for the correlation ID added to the VRU transfer label - change to

incoming called-number 9999999999T

Regards,

Geoff

[Luis Yrigoyen]

Geoff,

     Thanks for the quick reply.

What I want to do is forward the call to an outside number (i.e. a cell phone) if CVP can not be contacted.

I'm using "after-hours-agent0" as per the guide: "If you don't care about time-of-day routing, and simply want a last-resort transfer mechanism, put one of more DNIS's in the after-hours-agent slots and don't define any times.Any failed call will always be directed to the list of after-hours-agents."

To test it I shut down interface so that the gateway can't get to the call server (or any network).  The result I get is that the call gets answered by the gateway but there's dead air and it disconnects after about 1 minute.

I made the following changes to the config but still doesn't work.  I'm following instructions from the admin/config guide and a cvp book but I just can't get it to work.  Couldn't it be because I'm using FXOs?

thanks

service survive flash:survivability.tcl

  param alert-timeout 20

  param after-hours-agent0 3051234567

  paramspace callfeature med-inact-det enable

  param setup-timeout 7

dial-peer voice 99 voip

description For Incoming Leg (Type 10 label and Correlation ID)

service bootstrap

incoming called-number 99T

dtmf-relay rtp-nte h245-signal h245-alphanumeric

codec g711ulaw

no vad

dial-peer voice 49302 pots

service survive

incoming called-number 4930

direct-inward-dial

[geoff]

can you test it by inserting an agent's phone number as the after-hours-agent0 just to see if it works on the FXS port?

Regards,

Geoff

[Luis Yrigoyen]

I'm not sure if this what you mean; but this is what I did and still didn't work.  Thanks.

To test it I shutdown the interface.

I have a call manager in this lab (x.x.x.81) and I have one of the FXOs lines pointing to the call manager (translation rule)

I don't have SRST configured.

voice-port 0/0/0

trunk-group FXO-Ports 1

supervisory disconnect dualtone mid-call

output attenuation -3

no echo-cancel enable

no non-linear

no vad

playout-delay maximum 250

playout-delay nominal 200

playout-delay minimum high

playout-delay mode fixed

timeouts call-disconnect 5

timeouts wait-release 5

connection plar 4198

description DID 305-826-4198

dial-peer voice 4198 voip

description DID 305-826-4198

destination-pattern 4198

session target ipv4:x.x.x.81

voice-class codec 1 

voice-class h323 1

dtmf-relay h245-alphanumeric

no vad

!

I change the survive service to dial 4198 in case of outage

service survive flash:survivability.tcl

  param alert-timeout 20

  param after-hours-agent0 4198

  paramspace callfeature med-inact-det enable

  param setup-timeout 7

[Luis Yrigoyen]

I forgot to mention that when I enter the param after-hours-agent0 4198 under the "service survive flash:survivability.tcl"

I get the following message:

Warning: parameter after-hours-agent0 has not been registered under survive namespace

thanks

[Luis Yrigoyen]

I finally got it to work.  I first added a reference to the voice port on the pots dial peer (see below) but that still didn't do it but then I read on another post that their survivability was kicking in after 1 minute so I decided to wait and it was on.

Now all I need to figure out is why the dalay,I think it has to do with a bug.

thanks

dial-peer voice 49302 pots

service survive

incoming called-number 4930

direct-inward-dial

port 0/0/1  (it also works with trunk-group-label source "name of trunk")

[Luis Yrigoyen]

It's a bug: CSCth50439    survivability tcl script takes 60 seconds to kick in

I opened a TAC case and the engineer found it right away.

*********************************************

Symptom:

In a Cisco CVP solution installation it might be observed that the survivability tcl script takes 60 seconds to kick in on the gateway when the CVP server is down or unreachable.

Conditions:

When the CVP server is down or unreachable. So far seen in CVP versions 4.1 and 7.0.2.

Workaround:

comment out the following line in survivability.tcl:

# set fsm(any_state,ev_proceeding)              "StopTimer same_state"

and load the new file and service on the gateway, after that you can set the following under the survivability service to desired time in seconds to control the  time it takes to start the script.

    param setup-timeout 

************************************************

thanks again for all the help.