Failed to get the ICM RONA with CVP working

Answered Question
Jul 10th, 2010

Hi,

Currently we are in the midst of implementing the UCCE 7.5(8) system with CVP 7.0(2). The Unified CCM is 7.X.

We have encountered some problem with the RONA call with the contact center call flow in CVP. I understand that the RONA configuration for CVP is different from the IP IVR. The Ring no Answer Timeout in ICM Agent Desk Setting does not really redirect the "Ring no answer call" from the first agent to the next one or back to the CVP call studio script.

I also come acrosss that the the RONA is being controlled by the Router Requery feature in CVP. Even when I have enable the "Target Requery" in the Queue to Skill Group node, the phone of the first agent would ring for 60 seconds before the next step of the call flow takes place.

At the moment, I have stopped and disabled the CVP Voice Browser service and everything runs fine except the RONA. Must I start the Voice Browser service in order for the RONA feature to work? Whenever I start the Voice Browser service, it always says that the VB is unable to connect to the Gatekeeper, but we do not have any gatekeeper in our implementation. I also noticed that the service causes the error message to be played back when the dialed number is called after a while.

Appreciate any advice on how to set up the RONA feature in CVP.

Thanks & Regards,
Eric

I have this problem too.
0 votes

CVP RONA with SIP is quite easy to configure. Let's take this through in logical steps.

1. Forget the Voice Browser (H323 service). Leave that disabled. Forget about agent desk settings for a while - we'll come back to that. Assume we have just one agent in the skill group we are testing.

2. The 60 seconds comes from your gateway. In the sip-ua section you probably have something like this:

sip-ua

  retry invite 1

  retry response 1

  retry bye 1

  retry cancel 1

  timers expires 60000

  sip-server ipv4:a.b.c.d:5060

  reason-header override

The timers expires is 60000 millisecs, or 60s. When that timer goes off, ICM cancels the call at the ringing phone, and unless you have router requery configured at the queue to skill group node, the call is terminated and it falls back to survivability and you hear the error message. 60000 is the minimum value of this timer, but as an experiment, increase it to 90000 and you will see that it takes 90 secs before the ringing phone is terminated.

3. Add Router Requery to your queue to skill group. Now at the 60s mark, the call will come out the "X" port. You must take this and redirect it back into the queue to skill group. Use a line connector to make it obvious what's going on. Since other error conditions could also make the call come out the "X" node, insert an IF node. Test the value of Call.RequeryStatus. If greater than zero, this is a requery call, so run it around and back into the queue to skill group. If it's not greater than zero, it's some other error condition which you handle in the normal way.

Now make the call. At the 60s mark the call will disappear from your agent but come back within a second. Requery is working. Check?

4. When a call is put back into the queue like this, and no one is ready, the conventional approach is to increase the priority of this call so it jumps to the front of other queued calls. The way to do that is to add an IF test on Call.RequeryStatus after the check port. If this is greater than zero, insert a Queue Priority node and lower the value (higher priority). Then go to your music loop as normal. If you have a greeting for normal calls, the requery path would bypass this and go straight to the start of the music loop.

Make the call again. Just before the 60s mark, make the agent not ready. Now when requery happens, the call will disappear from the agent and go back into the queue and the music starts again. Look at the script in monitor mode to check that it is following the correct path.

5. Now to adjust the timer. Open your Ops Console and go to the SIP page for your Call Server. In the "Dialed Number (DN) patterns" section, below the Send to Originator section there is the "Patterns for RNA timeout on outbound SIP calls". Add an entry to the Dialed Number (DN). Insert the AGENT EXTENSION. Put 15 in the Timeout window and press Add. You will see EXTENSION,15 in the list.

Save and Deploy.

Remember - the entry is not the DIALED NUMBER (DNIS) you associated with the Call Type to run your script. It's the agent's extension. You can use the period as a wild card for a single digit and the greater than (>) for 1 or more digits at the tail in the usual way.

Now make the test. At the 15s mark the call will be pulled away from the ringing phone and requery will run. Check.

6. Finally, agent desk settings. Set the "Ring no answer time" to 10s. It must be less than the SIP timer by a few seconds. The agent MUST be made not ready by ICM before SIP takes the call away, or you open the door for other queued calls.

Now make the test. After 10s or so, the agent is made not ready and shortly after, the ringing phone stops and the call is back in the queue. Don't make the timers too close - 5s difference is about right.

Regards,
Geoff
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (3 ratings)
Loading.
Correct Answer

CVP RONA with SIP is quite easy to configure. Let's take this through in logical steps.

1. Forget the Voice Browser (H323 service). Leave that disabled. Forget about agent desk settings for a while - we'll come back to that. Assume we have just one agent in the skill group we are testing.

2. The 60 seconds comes from your gateway. In the sip-ua section you probably have something like this:

sip-ua

  retry invite 1

  retry response 1

  retry bye 1

  retry cancel 1

  timers expires 60000

  sip-server ipv4:a.b.c.d:5060

  reason-header override

The timers expires is 60000 millisecs, or 60s. When that timer goes off, ICM cancels the call at the ringing phone, and unless you have router requery configured at the queue to skill group node, the call is terminated and it falls back to survivability and you hear the error message. 60000 is the minimum value of this timer, but as an experiment, increase it to 90000 and you will see that it takes 90 secs before the ringing phone is terminated.

3. Add Router Requery to your queue to skill group. Now at the 60s mark, the call will come out the "X" port. You must take this and redirect it back into the queue to skill group. Use a line connector to make it obvious what's going on. Since other error conditions could also make the call come out the "X" node, insert an IF node. Test the value of Call.RequeryStatus. If greater than zero, this is a requery call, so run it around and back into the queue to skill group. If it's not greater than zero, it's some other error condition which you handle in the normal way.

Now make the call. At the 60s mark the call will disappear from your agent but come back within a second. Requery is working. Check?

4. When a call is put back into the queue like this, and no one is ready, the conventional approach is to increase the priority of this call so it jumps to the front of other queued calls. The way to do that is to add an IF test on Call.RequeryStatus after the check port. If this is greater than zero, insert a Queue Priority node and lower the value (higher priority). Then go to your music loop as normal. If you have a greeting for normal calls, the requery path would bypass this and go straight to the start of the music loop.

Make the call again. Just before the 60s mark, make the agent not ready. Now when requery happens, the call will disappear from the agent and go back into the queue and the music starts again. Look at the script in monitor mode to check that it is following the correct path.

5. Now to adjust the timer. Open your Ops Console and go to the SIP page for your Call Server. In the "Dialed Number (DN) patterns" section, below the Send to Originator section there is the "Patterns for RNA timeout on outbound SIP calls". Add an entry to the Dialed Number (DN). Insert the AGENT EXTENSION. Put 15 in the Timeout window and press Add. You will see EXTENSION,15 in the list.

Save and Deploy.

Remember - the entry is not the DIALED NUMBER (DNIS) you associated with the Call Type to run your script. It's the agent's extension. You can use the period as a wild card for a single digit and the greater than (>) for 1 or more digits at the tail in the usual way.

Now make the test. At the 15s mark the call will be pulled away from the ringing phone and requery will run. Check.

6. Finally, agent desk settings. Set the "Ring no answer time" to 10s. It must be less than the SIP timer by a few seconds. The agent MUST be made not ready by ICM before SIP takes the call away, or you open the door for other queued calls.

Now make the test. After 10s or so, the agent is made not ready and shortly after, the ringing phone stops and the call is back in the queue. Don't make the timers too close - 5s difference is about right.

Regards,
Geoff
eric.neoh Mon, 07/12/2010 - 04:25

/* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

Hi, Geoff.

I have tried the steps as advised in your previous email, and the RONA is working now! Thank you so much for the details of setting up the RONA with CVP. Really appreciate it. However, whenever the call has "rang-no-answer" on the first agent's phone, I saw a CTI warning message saying that "The call has failed with a cause value of 27". The call could be routed to another available agent or back to the CVP queue without any problem.

So what does the error message mean and how do we solve it?

Thanks & Regards,
Eric

Luis Yrigoyen Mon, 11/15/2010 - 14:49

Eric,

     Did you ever find out what the "The call has failed with a cause value of 27" error was all about.

We're running ICME 7.5(9) and CVP 7.0(2),  Our RONA settings are still pointing to a Dialed Number terminating into a RNA script with all the skill groups nodes in it.

Im in the process of correcting it following Geoff's instructions but I noticed that today an agent reported the "The call has failed with a cause value of 27" error.

Does it have to do with RONA misconfiguration?

thanks

eric.neoh Mon, 11/15/2010 - 15:20

Hi, Luis.  This is a non critical "cosmetic" warning message. You can get rid of it by applying the Ctios 7.5.9 patch on the ctios server as well as the agent desktop.  Regards, Eric

Luis Yrigoyen Tue, 11/16/2010 - 07:24

Eric,

     Thanks for the quick reply.

I was wondering what effects/symptoms would have/show when the RONA is misconfigured under a CVP enviroment?

My current RONA config:

CVP:

No CVP timer or extensions configured under the "Dialed Number (DN) patterns" "Patterns for RNA timeout on outbound SIP calls"

ICM:

Target Requery Disabled in the Skill Group nodes

ICM script called RNA with all a bunch of Route to Skill Group nodes referencing every skill groups.

DIALED NUMBER called RNA under FL_CVPCS_PIM.rc.RNA

CALL TYPE called RNA assoiciated with the DIALED NUMBER and terminating into the RNA script

Desk Settings:

Ring no answer time:             16 secs

Ring no answer dialedd number:    FL_CVPCS_PIM.rc.RNA

Logout non-activity time:         20 secs

CM:

Forward No Answer Timer:        20 secs

Call flow:

The calls are originating from the PSTN. We send the calls directly from the gateway to CVP without invoking UCM.

All the calls first hit a CVP script before getting to the Skill Group node.


thanks

2. The 60 seconds comes from your gateway. In the sip-ua section you probably have something like this:

sip-ua

  retry invite 1

  retry response 1

  retry bye 1

  retry cancel 1

  timers expires 60000

  sip-server ipv4:a.b.c.d:5060

  reason-header override

The timers expires is 60000 millisecs, or 60s. When that timer goes off, ICM cancels the call at the ringing phone, and unless you have router requery configured at the queue to skill group node, the call is terminated and it falls back to survivability and you hear the error message. 60000 is the minimum value of this timer, but as an experiment, increase it to 90000 and you will see that it takes 90 secs before the ringing phone is terminated.

Stupid to make this comment without validating. In fact, that value applies to SIP messages from the User Agent on the gateway, but not the one that finds the agent. That's done by the CVP SIP Back-to-Back User Agent (b2bua). Changing the gateway has no effect - it still times out at 60s.

If you look in the Cisco\CVP\conf\sip.properties file (pushed to the Call Server by OAMP) we see this (when no patterns are pushed):

#DN patterns to match for setting call specific Expires header (RONA) on outbound invites from sip b2bua

#format = DN pattern, timeout in seconds (min 5, max 64000,default 360)

SIP.OutboundInviteTimeout=

It's clear that the B2BUA has the timer, and it has a default value. The only thing is, it's not 6 minutes!!!

I'm guessing the comment is wrong.

Sorry to mislead anyone.

Regards,
Geoff
eric.neoh Thu, 07/15/2010 - 03:57

Hi, Geoff.

Now we have encountered an interesting issue, the RONA is working now actually. When I tried to call the hotline, the call will ring the agent's phone, however if the call is not answered within the RNA time, sometimes the caller will heard a short silent tone and subsequently the call is ended. By right, the call in queue message should kick in. Based on your experience, do you have any idea of this kind of problem?

Thanks & Regards,
Eric

I don't think I can really help.

I have it installed on my lab machine.

When it is set up normally, the call RONAs, the agent is made not ready, I hear the queue music again, press ready, and it runs the same cycle. I did not have any issues in the multiple tests I made of this.

If I remove the agent desk setting 10 second "not ready" timer and just make a call to a single ready agent, after a couple of rings, CVP pulls the call away but it requeries and finds the same agent, and so it rings again, and is then taken away. And so on. I had this running for 10 mins or so and it always worked.

I do have a customer who is experimenting with RONA (not in production) and they can get it to play up. What they see in the routing script is the call cycling through the X port of the queue to skill group node, and going back into the queue to skill group, back out the X port and so on until it exceeds the max number of nodes (1000). I can't duplicate this error, but I've seen the script in monitor mode. I don't know the cause.

Put your script in monitor mode and check.

Regards,

Geoff

Edward Umansky Thu, 07/15/2010 - 12:51

Make sure you restore your sip-ua settings and use the second method geoff recommended for setting the RONA timeout, i.e. using the OAMP console to add a DN pattern and timeout. Also make sure the timeout in your agent desk settings is less than the timeout you set in CVP. Also make sure that if there is a UCM no answer timeout that it is longer than either the one in ICM or CVP.

Are the calls you are testing with originating from an IP phone or PSTN? If from the PSTN, are you sending the calls directly from the gateway to CVP without invoking UCM (you should be). If the call is UCM originated, make sure you do a Send to VRU before routing a call with router requery, UCM doesn't support router requery.

In general double check the settings on your gateway and SIP trunk in UCM, make sure they match what is documented in the CVP config guide. I have a few customers using CVP with SIP and RONA always worked fine.

Also make sure that if there is a UCM no answer timeout that it is longer than either the one in ICM or CVP.

Good suggestion.

I forgot to mention that, since I always have no call waiting, no voice mail, no "no answer timer" on the Contact Center line. If they need voice mail, call waiting etc. then they also have a private line.

Regards,

Geoff

Luis Yrigoyen Wed, 11/17/2010 - 08:26

I made the changes to configure CVP RONA and everything worked fine regarding the RONA and the target requering but now calls for non-agents going to voicemail hang up.

So what I've done is configure it only for 2 extensions (7672 and 7122) for testing and this is what I have:

CVP:

"Dialed Number (DN) patterns" "Patterns for RNA timeout on outbound SIP calls"

7672,16

7122,16

ICM:

Target requery enabled

     Desk Settings:

     Ring no answer time:                  16 secs

     Ring no answer dialedd number: 

     Logout non-activity time:              20 secs

CM:

Forward No Answer Timer:        20 secs

We have two types of Inbound calls, calls that go to skill groups and calls that go to non-agents.

1. Call flow for Skill Group calls:

The calls are originating from the PSTN. We send the calls directly from the gateway to CVP without invoking UCM.

All the calls first hit a CVP script before getting to the Skill Group node.

2. Call flow for non-agents:

The calls are originating from the PSTN. We send the calls directly from the gateway to CVP without invoking UCM.

All the calls first hit a CVP script here callers enter the extension number.  ICM then returns a label with the extension number entered.

Under 1, if I login as agent using 7672, the RONA works fine, if 7672 is no answer it rings on 7122, all fine.

Under 2, NOT logged as an agent, the phone rings and if is a no answer it just hangs up.

If I remove the 7672 from "Dialed Number (DN) patterns" "Patterns for RNA timeout on outbound SIP calls", the call forwards fine to voicemail.

Maybe I'm missing a route in the call server? or in the gateway?

Thanks

Eric Alisson Wed, 08/15/2012 - 05:56

Hi everyone,

My name is Eric and I've already done the RONA configuration using CVP.

But I have one question to do. The "Patterns for RNA timeout on outbound SIP calls" at CVP is a global configuration?

Because I have multiples operations using the UCCE and for exemple, I would like to agents at "operation Burger King" have 15 seconds ringing before the RONA works, and I would like to agents at "operation MacDonalds" have 25 seconds ringing before the RONA works.

Can I segmental the RONA configuration at "Patterns for RNA timeout on outbound SIP calls" for each agent?

vinodkewate Thu, 08/16/2012 - 12:19

Hi Luis,

Do you have Target Requery enabled on the Label Node?

If not, you have to enable it and then you have to connect that true branch of that Label Node to provide alternate treatment.

Regards,

Vinod

vinodkewate Thu, 08/16/2012 - 12:26

Hi Luis,

Do you have Target Requery enabled on the Label Node?

If not, you have to enable it and then you have to connect that true branch of that Label Node to provide alternate treatment.

Regards,

Vinod

Matthew Hammitt Tue, 06/04/2013 - 11:49

Just to clarify, has everyone here found that 60 seconds is the max configurable limit? I have a customer that would prefer a longer ring time for certain business cases but all roads so far have led to the same conclusion.

Regards,

Matt

Edward Umansky Tue, 06/04/2013 - 18:43

No, 60 seconds is just the default. You can change it to the amount you prefer by setting the RNA timeout pattern in the OPS Console. See Geoff's accepted answer above.

Edward Umansky Tue, 06/04/2013 - 19:08

Nevermind I see your other posts about the 60 second limit. Yes it looks like there is a 60 second limit here. What you can do to mitigate is resend the call to the destination label after requery expires and keep track of how many times you did that. Otherwise you can try to mess directly with the CVP properties files that Geoff mentioned earlier, but that is probably a bad idea.

Matthew Hammitt Wed, 06/05/2013 - 07:09

Update:

Apparently the CVP Call Server GUI in Ops Console is limited to 60 seconds as a maximum value and anything entered greater than 60s will default back to 60. If you need a longer RONA duration you will have to enter the pattern and timeout setting directly in the "SIP.OutboundInviteTimeout=" portion of the "sip.properties" file on each Call Server, indicated by Geoff above.

Also be careful if you do this as it's possible that future 'Save & Deploy' commands through the GUI will overwrite your custom configured sip.properties settings with the GUI max of 60s.

Regards,

Matt

carlos osorio Wed, 07/10/2013 - 15:40

Hello guys, maybe this is a dumb question but how will this behaviour be affected if its configured for RONA but agents have auto answer enabled?

I´m having some troubles with error code 50010 getting back agents to NOT READY state and everything points to RONA configuration evena Q850 cause 19.

Please help!.

Thanks.

CAOSORIO.

Actions

This Discussion