1Paul Withers  11/21/2012 6:00:17 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Sounds very interesting. Declan's post on ValueChangeListeners was very useful for me for audit trails, although it needed a couple of tweaks.

Are you using the Extension Library for integrating via JDBC?

2ChrisJC  11/21/2012 6:05:43 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Paul,

Thanks - Decs post was of real help initially.

Re "Extension Library" - I am kind of but with a little twist. Its all done with Java but certainly utilises some of the ExtLib stuff. Hope that makes sense (without going into the nitty gritty right now)...

3Jesper Kiaer  11/21/2012 6:18:12 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Looking forward to it!

I have a customer with a similar scenario, so I am very interested to hear your solutions.

4ChrisJC  11/21/2012 6:29:54 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Jesper,

Sure - there is so much to cover in this it may take a bit of time - keen to do it right. Hopefully looking to build up some resources around blog postings, videos, presentations etc. The issue around views / reporting / volumes is an age old one.

However there is a lot around the use of Java throughout the application too that I think may be useful - it does follow on nicely from a presentation I did at LS12.

Again - if there are specifics you would like to focus in on then feel free to ask....

5Nick Wall  11/21/2012 6:43:13 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Sounds great. I will follow your posts with great interest.

We have a web application that has been up for a while, we generate a LOT of audit data (every single field change, every button press...it's a requirement), which we store in SQL, we get and retrieve this via web services, we batch the data across. We also use a SQL data store for some aspects of the app, accessed via beans and an MVC model (so we can switch between a Domino or SQL data layer), but we are not doing it efficiently. At the moment, the JDBC driver is in the DB, so it opens a connection each time it is accessed, we need to switch this to connection pooling.

I have a test server set up where I tried to set up an OSGI MySQL plugin, but I can't get it to work. I followed chapter 12 of XPages Extension Library Book, but I think I need to try again using DB2 as in their example, there are a lot of steps and my IDE does slightly diff things to book and I guess I am doing something incorrectly. Anyway, would be very interested to hear how you deal with JDBC drivers, connection pooling, clustered environment etc.

If I get my OSGI plugin working (someone had one posted online but had to pull it due to some legal issue) I will post my results. I've just seen this, so need to check out when I have time: { Link }

6ChrisJC  11/21/2012 6:47:19 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Nick,

Thanks - Great comments.

I setup the OSGI plugin with the MSSQL driver and it works very well - from memory I followed the DB2 steps. Getting the right driver is key I guess.

Maybe as part of the process I should revise and document this configuration. It will at worst act as a reference for me in the future in doing it..

Keen to chat around "optimisation" as this is something there is not a lot of information around in XPages land.

7Baiju Thomas  11/21/2012 8:20:09 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Nick,

I could setup SQL Server jdbc driver as an OSGi plugin just by following the DB2 steps... It worked the first time. It is a lot of steps, but you just need to follow the documentation...

I am working on an application which uses SQL Server as backend and XPages as the front end. I am using the new JDBC feature in extension library. My approach is to define the JDBCQuery and JDBCRowset for the datasources and once you set them, you could just work with them as you would with an NSF backend. I am trying to minimize the differences beyond this point. So all the validations, conversions ,simple actions etc. will work just the same way as in a regular XPages application with NSF backend.. The only place the code will change is where you set the dataSources. Once the data source is set, it is a bean and you can work with that data just like a notes document or notes view. It is all working great so far, but I have a long way to go before I complete this application.

Upto this point, it all worked perfectly as I had hoped and performance has been great..

8ChrisJC  11/21/2012 8:26:26 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Baiju,

My app is sort of similar. I have data access classes which read lists of records that relate to a main document in Notes (sort of one to many). All the docs on save are written to SQL. On open of the doc again the "many" records are read into an arraylist for display and maniplution.

All the converters and validators etc dont care about the data storage (as that all happens on save - after conversion and validation has passed).

9Gayathri Viswanathan  11/22/2012 3:22:45 AM  XPages Case Study Application - Procurement Tracking - A Hybrid NSF / SQL Application

Chris,

This would be a great addition to the XPages resources that are being built up. I am working with couple of customers where we are creating POC to convince them that they can leverage their existing domino infrastructure far beyond what they have done before and showing them integrated UI where we have data coming in from SQL has been a good value add. Looking forward to your series to learn more best practices