AnsweredAssumed Answered

Caching lookup results in process properties?

Question asked by TehNrd on Mar 20, 2012
Latest reply on Jun 8, 2016 by BhoopathireddyNaradi3621
When building a process one of the most, if not the most common step, is checking to see if a record already exists in the target system. Then proceed down the correct path based on this result. With SaaS apps like this is a painfully slow process. The biggest issue is that Boomi does this check one record at a time, see this idea to improve that: .
I have been thinking about how to improve this and I have an idea but I want to make sure I am not crazy or this will cause serious performance issues.
Rather than checking for the existence of a record with a connector lookup one at a time, what about pulling the entire target table from Then populate a bunch of process property values with the key being the unique identifier between systems and the value being sfdc record Id.
This approach would be way faster. A sub process could pull down the records 200 at a time and then place the results in process properties. In the next branch you could use a simple decision shape to check for the existence of the process property and route the record appropriately.
The one big question mark around this is the amount of process properties. If I am replicated a sfdc object this could be 50K+ properties.
Will this work? Am I crazy?