💡 Oracle EPM July Update – Smart Split Is Here! A Game Changer for Data Loads

I was going through the Oracle EPM July updates and one feature that truly caught my eye is Smart Split – and I must say, this one’s a game changer!

Let’s first unpack what it is before diving into a real-world scenario that many of us have found painfully familiar.

Oracle has done a great job documenting how Smart Split works and what to consider when using it (you can read the full details here), so I won’t go deep into the technicalities. But in a nutshell:

Smart Split is a new feature within Data Integration that helps when importing a large volume of data exceeding Essbase governor limits. It automatically splits the load into smaller data slices—without needing to create multiple integrations manually.

Yes, that’s right—no more manually creating separate data load rules just to avoid QUERYRESULTLIMIT errors!

Now, this immediately reminded me of a recurring issue I faced in past with Oracle Fusion GL loads when integrating actuals. Anyone else remember this dreaded message?

 “Exceeds the limit [1000000] set by the QUERYRESULTLIMIT”

It looks something like this in the error logs:

ERROR from the Process log=================

INFO [AIF]: ExtractBalanceData:Error::Cannot perform cube view operation. Essbase Error(1200766): The number of query result cells exceeds the limit [1000000] set by the QUERYRESULTLIMIT configuration setting

So frustrating! Everything would look perfect during SIT or even UAT or post Go Live, but once the ledger starts to scale, the interface crashes due to the Essbase governor limits. And then begins the scramble—splitting loads by Account, Entity, or Cost Centre to make it work. Not fun.

Well, Smart Split could finally put an end to this madness. By allowing us to define logical splits within the integration itself, it’s set to significantly reduce overhead and make our integrations far more resilient as the volume grows.

I’m really excited to explore this once it goes live this Friday. I’ll definitely update this thread with anything new or clever I discover during my testing. Got some other use cases in your mind? Feel free to drop them in the comments!

Until then…

Happy EPM-ing! 💙