CIs: Connecting to PeopleSoft Jolt sever from VB.net. Why & How. Part 1

[If you are looking to Load 100k+ rows into PeopleSoft quickly, this isn't the post for you! This is a multi post piece on Component Interfaces. Hopefully it is informative reading...]

Part 1:

There have been a lot of times when we have needed to mass load some sort of data. One of the main issues as always been load directly into PeopleSoft DB tables is just yucky... :) Some times unavoidable, but almost always yucky... I guess i should explain what i mean by yucky... The clearest example is person bio demo data. When you use 'add/update a person' to add a new person the emplid is give to you, the person's data is strewn over a plethora of tables and all the effective dating stuff is taken care of for you. Not to even touch on search match possibilities; cause we all love dupes so much...

Moving data around between platforms is always a headache. Then there is the migraine called synchronizing data... anyway, say you have an application that really needs data moved from it to Oracle's PeopleSoft. You could write something to connect to both DB and just pass data back and forth, but you want a solution that is more flexible, able to be used by an end user (securing DB level access for end users, do i need to say more...) and you want to be able to leverage data validation from the application tier. Sounds great. You have two options: CI or Web service.

As of 8.48, web services are a good option, but they are still lacking some (this could totally be from a lack of understanding from my part) and are a little ... um.. clunky?... when it comes to dealing with re-posts and data validation. They are great for posting small data objects from remote systems that you can almost be curtain the data is always good.

I have come to really have an appreciation for CIs. At first I really didn't like them because of the fact that they are based on old COM objects, don't work too nicely with .Net (I now have some tricks... :), setting up the environment was a pain, maintaining distributions was painful too and they are inherently SLOW (slow-as-a-wet-week). After I overcame these little hurdles, or learnt to live with them, the results are quite workable.