Queuing Emails from EIM in UCCE

Unanswered Question
Feb 13th, 2008
User Badges:

Hi,

i have an EIM 4.2 running with UCCE 7.2.3. When the route request reaches the ICM script the email is transferred to the LAA agent - which is good. When no agent is logged on. the email is not queued until an agent is logging on instead the email is sent to the execption queue in EIM and the email is lost for the agents.

Does anybody has an idea how to solve this problem?

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (5 ratings)
Loading.
ake Thu, 02/14/2008 - 04:12
User Badges:

Hi, why don't you use the queue to skillkroup node, that should help. LAA don't queue anything, it just tries to find an available agent.

alex.dersch Fri, 02/15/2008 - 02:46
User Badges:

Hi Ake, thank your for reply. Unfortunately it did not work with the queue to skill group node. The route request terminates in the MR PG with a MR_peripheral::Send_ToApp_NewTaskFailureEvent. I attached the dumplog for the pim.


Regards


Alex



Attachment: 
antonio.vazquez Fri, 02/15/2008 - 02:57
User Badges:
  • Bronze, 100 points or more

Hi Alex,

could you send a screenshot with the script you are using?


Regards,

Antonio

antonio.vazquez Fri, 02/15/2008 - 03:34
User Badges:
  • Bronze, 100 points or more

Hi,

I have viewed your ICM script. IM is similar an IVR system, so I think you need a Run External Script node in your script as you use it in chat integration.


Regards,

Antonio

alex.dersch Fri, 02/15/2008 - 03:51
User Badges:

Hi Antonio,


i followed the cisco installation guide to create the script. They don't mention to use an external script for queuing. I attached the sample script from the Inst Guide.


Regards


Alex



Attachment: 

Alex, your script is fine, although I have the exit from the wait 3600 lead to another wait 3600 and then loop back to the original wait - an infinite loop. Otherwise, the router will terminate the route request after one hour. This is just a nicety, though - your script should work.


Regards,

Geoff

jeroen1964 Mon, 02/25/2008 - 05:26
User Badges:

Hi Alex, i run into the same problem using this example script. I did found a TAC case that describes how to increase the max wait time in the script.

I also found out that when an agent is configured in CIM as ICM agent, you cannot assign local queues to that agent anymore...

Did you find a solution in the meantime?

Jeroen

Chad Stachowicz Mon, 07/07/2008 - 14:47
User Badges:
  • Silver, 250 points or more

I have a question. No Matter what workflow I setup. The incoming request to ICM when an email is recieved is using the default Call Type for the MRD. This means that every email regardless of which address it comes in from. Is going through the same ICM script then into the same queue. In EIM I see it has a place for the script selector to run per Queue. Am I missing something, something misconfigured in ICM?


Cheers!


Chad

Chad,


You have followed the eGain instructions too closely. They don't really understand ICM in the way they set the call type on the MRD. No one does that.


Anyway, there is a solution. Create a new call type, create a new script selector on the EIM MRD; in EIM, create a new queue and select the PG, so it's an integrated queue. Now you can choose the script selector.


Clearly, different email aliases will map to different script selectors (through the integrated queue) and you can do this in the one workflow with the branch node.


Now you can have a single ICM script that detects the Dialed Number / Script Selector and queues to the appropriate skill group.


Regards,

Geoff


Chad Stachowicz Mon, 07/07/2008 - 16:24
User Badges:
  • Silver, 250 points or more

Geoff,


This is what I dont understand is how the script in ICM should look. Would you mind dumbing down the procedure to a really low and stupid level for me? This is a very urgent matter needed by the morning. Maybe a screenshot? Also I made an incoming workflow for each email addy. then a queue node after that..


Cheers,


Chad

Chad,


The ICM script could not be any easier. It's basically "Queue to Skill Group" and a Wait node. There are some niceties, but you can skip them if you want to just see it work.


1. Check the value of the Router registry setting - MaxTimeInQueue. It's under RouterA\Router\CurrentVersion\Queuing. The default is 3600 secs. You don't want this to expire first as it will push the email into the Exception Queue on the eGain side. Let's assume you leave it alone.


2. Make the Wait node say 3590 secs. Make the exit of the Wait node go to a label that returns the special label on the MR PG routing client to push the email to the non-IPTA skill group. This is of the form: "LBL_".


The enterprise skill group with the name above does NOT have "ICM picks the agent" checked. Make sure the enterprise name is exact (ICM will want to prepend PG and MRD - but you can change it).


3. Failure from the Queue to Skill Group node should also return this label.


If you don't want to have a dedicated agent who is ONLY in the non-IPTA group you can skip all this and let it go to an End node - the email will never be lost but will be in the Exception Queue on EIM.


4. In your case, use a number of DN nodes to partition the calls to send them to the appropriate skill groups. I assume all the script selectors are associated with the same call type and this is mapped to your scheduled script.


5. The best trace level to set on the MR PIM is leave the trace mask alone and to configure UserData in the registry and add "df". This will show the task request and responses with the most useful information.


6. The skill group(s) you queue to are, of course, those that you have integrated into EIM. And your integrated agents are members.


7. Do not have an agent in a IPTA skill group and a non-IPTA skill group - the router will never deliver an activity if you do that.


Really, there are lots of stars to align to get EIM to work, but the scripting is the most trivial part.


Regards,

Geoff


Chad Stachowicz Mon, 07/07/2008 - 17:32
User Badges:
  • Silver, 250 points or more

Geoff,


I am trying to make an IPTA. I have attached the screenshots of some relevant configuration. I am unsure why I cannot trigger script selector's :(. Remember I mentioned that if I set a default call type for the MRD the script always runs just fine and emails are all routed to that script. However I feel from what I have read and how things import, I should be able to run any script seector from seperate workflows in egain.. Right now all I want to see is my script increment, as emails enter it so I know things are working correctly between ICM and eGain. If I have a misunderstanding and need to do non-IPTA please let me know..


Chad



Thanks a bunch!


Chad



Attachment: 
Chad Stachowicz Mon, 07/07/2008 - 17:46
User Badges:
  • Silver, 250 points or more

oh that reminds me of another awesome question! How do I get stuff out of the exception queue to any agent that was to log onto the system? Or is this a pulled activity? I think I have plenty of emails sitting in my exception Queue..


Chad

Most users can see the Exception Queue - but only certain users can transfer out of it. Just run a search and nominate that the queue name contains "Exception".


The pa (Partition Administrator) can select items in the Exception Queue and transfer them to the input queue to push them through your workflow again.


Regards,

Geoff

Kapil Atrish Tue, 07/08/2008 - 08:58
User Badges:

Hi Geoff,


I am really suprised to know if PA can do it. Pl let me know how does PA see the contents of exception queue? I know PA can run the monitor and see how many e-mails but not sure if he can see the contents and reassign them also.


Thanks,

inner_silence

Kapil Atrish Tue, 07/08/2008 - 07:30
User Badges:

Hi,


Exception queue is a default queue of EIM. You need to create a standalone Agent on EIM i.e. the Agent doesn't exist in ICM and give this Agent pull permissions on Exception queue. That works beautifully for me. Let me check your other concerns and answer soon.


Thanks,

inner_silence

You can forget the non-IPTA stuff for a while. Is your skill group a IPTA skill group - that's the default, so I am sure it will be.


In your ICM script, use a Queue To Skill Group node leading to a Wait node. The way you have it, the call will only go to an agent who is available. It will not queue. The manual is wrong.


Turn on UserData trace on the MR PG as I noted above. When the task is sent up by the PG you will see a NEW TASK trace, and info on the script selector used will be obvious.


You can also use rttrace to increase tracing on your Call Router - set the usual check boxes and you will see the route request.


Regards,

Geoff


Kapil Atrish Tue, 07/08/2008 - 07:38
User Badges:

Hi All,


I've emails queued up for days in my integrated setup. You need to increase the registry in Router as rightly mentioned by Geoff, then you need the wait node in ICM script and lastly "Max duration" on either MRD or Media class for EIM. Lower value amongst these three will be in effect.



Thanks,

inner_silence

Kapil Atrish Tue, 07/08/2008 - 07:52
User Badges:

Hi Chad,


Mapping between ICM and EIM is something like following:


ICM_DN associated to ICM Call_Type.


When creating EIM_Queue select "Type IPCC/ICM and MRD "EIM" (or whatever name you've given). On the same scrren from the "Script Selector" drop-down choose ICM_DN as create above on ICM side.


Now, create you ICM_script and associate ICM Call_Type to this ICM_script.


After this configuration, any thing hitting EIM_Queue will hit ICM_DN >> ICM_Call_Type and finally ICM_Script.


Your EIM workflow would be something like:

Start (select aliases) >> Queue Node (EIM_Queue).


HTH.


thanks,

inner_silence



Actions

This Discussion