One way would be to remove the where clause and just add a StatisticsCalculator to the FeatureReader and select the attribute you want the max value for.  You have to read the entire dataset first to do this so keep that in mind.  

Using a where clause just filters rows - you can’t use an aggregate function like max in it.

SQLExecutor isn’t really a good option either since you’re reading from an SDE.

Good luck!

Hi, try this one: “update_start” = (SELECT MAX(“update_start”) FROM “database”.“table_name” WHERE “status” = ‘Success’)

I wasn't aware that this was possible, this post made me try it. 

This one didnt work but made me try a modification of it. The modified version worked:
update_start = (select max(update_start) from database.table where status = 'Success')

Thank you for the ideas!

Hehe yes I noticed my mistake, I have edited my reply but I was too slow it seems. Glad it worked out!

I just tried this too on my dataset and was able to get it work!  Very cool.  I get all the records that have the max value but that’s fine.

One HUGE thing to note is if you are querying against an SDE database (in my case with a MS SQL backend) make sure you are querying the correct table or view (in the WHERE clause in the FeatureReader)!

For example, if you have a feature class that has versioning enabled, use the _evw instead of the base table.