Smart Notifications Data Source Setup Recommendations -

Name	Points
Greg Moeller	4184
David Williams	3349
JonA	3288
Kat V	2984
Woozy	1973
Jimmy Chiu	1883
Kwane McNeal	1437
Ragu Raghavan	1348
Roger French	1311
mark.cook	1244

Michelle Sundermann

New Member

Posts: 3

12/2/2011 1:59 PM

We are in the process of rolling out Smart Notifications and have a question regarding setting standards for creating infosets. Our question for everyone has to do with the multiple types of data sources you can use within LBI and specifically Smart Notifications. That said, the one that is delivered when LBI is installed is the DME Query which the syntax is quite cumbersome we have found. However , one of the other options that we have been using is JDBC as our data source (which is essentially straight SQL).

So, now for the question, for those of you who have rolled out Smart Notifications, do you have a standard data source type you use for all your infosets? Wondering if most organizations are saying for all of our notifications we are going to use all DME Queries or all JDBC or one of the other choices?

Any advice or insight you can give would be greatly appreciated.

Jan

Veteran Member

Posts: 55

12/2/2011 2:18 PM

like you, I've found the DME to be very cumbersome although some others in my organization have used it - although not extensively. I much prefer the jdbc connection because I find it very easy to use and much more flexible. This isn't really a standard for us though. I guess we're still pretty new into this so we really haven't devleoped standards.

David Williams

Veteran Member

Posts: 1127

12/2/2011 3:03 PM

Using a DME is fairly standard but I find that if you need to join tables that Lawson doesn't have a relationship established for that a JDBC SQL query is the way to go.

Robby

Veteran Member

Posts: 87

12/2/2011 3:55 PM

JDBC gets my vote. I really like the flexibility of writing SQL to pull my infosets.

mark.cook

Veteran Member

Posts: 444

12/5/2011 12:01 PM

We have started rolling out the build of smartnotes in a DEV system to our super users. For them DME is very similar to Add-ins and makes a nice transition. If you are having IT build the SN then I don't think it would matter which one you used. It really comes down to comfort level. When you get to the JDBC, I know some people code differently, so my only thing would be to document what was done so if that person leaves or doesn't touch it for a long time they remember what was done and why.

Ruma Malhotra

Veteran Member

Posts: 412

12/5/2011 2:24 PM

If your infoset pulls a large amount of data jdbc source using sql is the best source. This avoids a lot of performance issue especially when the number of smart notifications in production may grow to a substantial amount in a period of time. This is also recommended by Lawson since dme goes through the extra layer of security checking as to whether the user writing the smart note has access to the data or not.

However if there are the "derived fields" involved in your smart note than there is no other option but to use dme.We've written smart noted with sql unless we have to absolutlely use dme.

Greg Moeller

Veteran Member

Posts: 1498

12/12/2011 8:01 PM

I disagree with you Ruma. Specifically your comment about 'If there are derived fields...'

Run

[b]derfldrpt productline[/b]

and what comes back should be what Lawson is using to arrive at those derived fields. You can use the same math in your sql queries.

Ruma Malhotra

Veteran Member

Posts: 412

1/4/2012 5:51 PM

My apologies. I am just getting to this post.

Completely agree to building the logic of derived fields in sql and that would resolve the issue. But sql is not everyone's cup of tea. There are some business users who build smart notes and are not very good in SQL.

For those users dme would be the only solution even though jdbc and SQL should be the first choice.