Hello

Here are the details on the DD Version, Application Hosting Environment, and database & JDBC driver details.

DD version 5.1.0.1101

Host Platform WebSphere Application Server V7 FixPack-21

Database product name : Microsoft SQL Server
Database product version : 10.00.2789
JDBC driver name : Microsoft SQL Server JDBC Driver 2.0
JDBC driver version : 2.0.1803.100

Our application has 10 dbop definitions each containing a single SQL operation of either a query or an insert. When we deploy the application to WebSphere and test we see that the connections are not released back to the connection pool at the completion of each dbop activity. Part of our flow is data-mining during call cleanup, and if we generate a large enough call foot-print we can exhaust the connection pool (30 connections) in a single call.

We need advice on why the connections are not being released. If you need any traces from WebSphere please detail what you need and I will provide.

Thanks

Thanks for your help.

We are setting up our datasource definitions, and just wanted to confirm that prepred statements are being used so we can correctly configure our prepared statement cache size.

We've used the DD database operation wizard to create our database operations files. Does the database support provided by DD use any form of prepared statements based on the SQL definitions provided using the DBOP wizard?

If the answer is no is there any possibility to implement the use of prepared statements through the DBOP support of DD?

Thanks

We've created some static VXML pages to invoke from the event handlers on the MPP. In these static pages we have used the <log> tag. Where on the MPP or VPMS would we see the logging generated by the <log> tag.

Ross

Avaya installed and configured the VPMS. What is on the page you describd is ::

User Name = <Default>

Password = This fiels is blacked out with dots.

What user ID and password is to be set here?


Is this inforamtion passed to the WAS to be used when the WAS attemps to send reportdata to the database?

Thanks

Ross

We've reviewed all of the points you've mentioned, and it appears we are configured correctly. We do have trace enabled, and we don't see any entries for the framework as you've described in trace.log although we do see the entry in report.log.

Not knowing what the sample rate is for the VPMS to fetch report data from the WAS we are waiting 10 - 15 minutes, and not seeing any of our report entries in the standard reports.

Do you have any other suggestions to determine if he data is sitting on the WAS waiting to be sent? I've checked the WAS logs and there are no reported issues relating to communications from the WAS to the VPMS.

Is there a way from the WAS or VPMS to force the retrieval of the report data from the WAS? Sort of a FLUSH all data.

Any suggestions would be appreciated.

Thanks

Ross

Thanks for the additional details. We are able to see the Report log being populated now (Which we will deactivate) but when I look in the trace logs of the call I see no log entries that the report interface was attempted. Should we see Report entries in the trace.log showing the attempt to send data to the VPMS?

We have waited 10 - 15 minutes looking for the Report entries to appear to the VPMS. Is there any delay in this information being sent to the VPMS?

Which port is used on the WAS to talk to VPMS? Is there any WAS configuration needed to ensure the WAS can talk to the VPMS?

Which log view on the VPMS will show any issues with the Report interface between the WAS and VPMS? I assume when the VPMS is notified that the call terminated then a query is made from the VPMS to the WAS to retrieve Report date.

Which Report view on the VPMS is the suggested view to see the Report data?

Thanks again for taking the time to help

Hello Ross

I appreciate your help. I understand your suggestion that we not activate report logging locally on our WAS. Just as an FYI we can't find a rrst.properties, and are curious where this property file is stored in our solution.

We have only configured our WAS to use the CTI connector to interact with AES. You've explained that the Report Item will result in writing to the VPMS database which is our objective. Since the Report Item we are using is placed in a datanode we don't see any VXML page fetched to the MPP for processing the Reporting Requst so I assume the interaction is between the WAS and VPMS. How is this interaction happening since we've not defined our VPMS location to the WAS.

Perhaps you an provide a few more low-level details on how the data is written to the VPMS database and where this communication link is defined. Again, we don't see any pages being fetched to the MPP so how is this information supposed to be written to the VPMS Database. We've run the VPMS reports and don't see our report entries.

Thanks

I have a 2 part question::

Part #1

We've added a Report Item to our callflow that collects a data item from he call. When we run this through the DD simulator we see the entry created in the report.log. We export the project and deploy it to our WAS 7.0 and run a test call. The report.log file is created, but the report entry generated by the application does not appear in the report.log file. Is there any issues with the report item processing on WAS?


Part #2

In addition to generating the WAS based report.log file entries we would like this information populated to the VPMS database as is describe in the Dev guide. What additional steps in the application, and/or in the VPMS configuration are needed to have this data stored on the VPMS so i can be utilized in VPMS reports?

Thanks

We're interestded in collecting some data from a call and storing it in the VPMS database using the Report Node. The Developers Guide for DD speaks about doing this suggests this note:

Note:
Note: If you are using an Avaya Voice Portal system,
you can view an automatically
generated Application Report of this log data.
See "Application Report" in your
Voice Portal documentation.

Which document has the "Applcaiton Report" section that we should review? Is it in the VPMS admin document, or is there another document for the MPP?

I've searched DevConnect and can't locate any documents on MPP. In addition to the question above is there any supporting technical documents on the MPP?

Thanks

Neil

IC is not an option for us. You've mentioned that I can use VXML under this failure scenario. From this point in the DD application how would we invoke VXML to execute a VXML transfer tag? Are you suggesting we invoke a static VXML page stored on the MPP? If so how can we invoke this VXML support from he DD app? Appreciate any advice you can provide.

As for using the CCXML option would we be able to invoke CCXML at this point in the DD application, or would the call need to have arrived via CCXML? I only see the ability to create a CCXML project which I assume would need to have been the entry point of the call. Appreciate any additional details for this approach

Neil

Of the two scenarios

#1 I understand that once we attempt a consult CTI
transfer and it fails we can only retry a consult CTI
transfer, but any other type of transfer is not
possible. In these situations we'll attempt to play
a prompt to the caller to try their call later. I think
we are set all set here.

#2) When the call starts and the interaction with the
the CTI connector fails to get CTI callinfo the
application then depends on the session data
for ANI & DNIS. For these calls the app will mark
the call not available for CTI transfer since the
initial interaction with the CTI connector failed.

I see two options to transfer he call:: The first is
a bridged transfer from an announce node, but this
holds the browser which is not an option.

The 2nd option is a Blind transfer from a datanode
which wants me to supply the callinfo which I
don't have since the initial interaction with the
AES failed.

Can you suggest which transfer I can use for this
failure scenario where the call was not able to
get cticallinfo at call start? The call needs to
transfer from the VXML and not rely on the
CTI connector.

Thanks for your help

I think I have a little different scenario that we can't seem to get working with a custom event. We want to create the follwing event processing::

When we receive either the first NM or NI we want to play a prompt we will call NMNI-1.

When we receive the 2nd NM or NI we don't want to play any prompt and just want to moved forward to transfer the call.

How do we prevent the NMNI-1 prompt from playing in the 2nd NM NI event processing.

We've looked at all of the code samples and can't get this to work.

Thanks for any help

Is the bridge/blind/consultation describing the use of 3 different nodes in that order to accomplish the MPP based transfer of the call?

Does this approach hold the app until the caller hangs up, or will the appliction regain control once the transfer completes or fails?

Are there any code examples of this transfer appraoch we can use as a guide.

Thanks

OK, that clears up my understanding.

I have on last scenario that I hope you can help with. A new call arrives, and the initial request is made to the WAS. The CTI connector encounters an error while retrieving the cticallinfo from the AES. The application catches the exception from the AES, and then uses the session to retrieve the ANI, DNIS, and UCID. The call continues on as normal. Now the application wants to transfer the caller, but is aware that the interaction with the CTI connector had an issue, and will not try a CTI transfer through the AES. The application wants to initiate a transfer from the MPP and this transfer does not have to attach data in this situation, but if that is possible then great. What transfer node should the application use to initiate a transfer from the MPP?

Thanks