We do, yes.
Do you run MRP? anything in the MRP queue?
Yes I’d clear out the MailQueue (there are 3 related tables MailQueueAttach and MailQueueBody) and only cause this is test… cough
2 Likes
Yes, we do.
Where do I see the MRP Queue?
Erp.MrpQueue
I would love to see the Event viewer on the W3WP if there’s any errors with a .NET error underneath it. Normally the errors come in threes. Follow Jose’s suggestion first, I’m troubleshooting from another angle.
There’s 47 records in the MrpQueue table, 11 in the MrpProc table, and 43 in the MrpProcQueue table.
Aaron, is your Enterprise Search setup as a continues process to update or do you have it on a schedule?
I had actually started there. There were no errors/warnings of note in the Event Viewer.
It rebuilds daily on a schedule.
So W3W doesn’t crash it just increases in size, correct?
Yup.
Are you running in 32 or 64 bit?
We’re on x64
This really sounds like a non-thread safe memory cache or contention/ deadlock issue not a memory leak.
We thought memory leak as it’s slowly eating more and more memory but that was just as assumption. 
Jose, wouldn’t thread traffic cause the same issue if their weren’t enough threads to handle the traffic?
Not much traffic, this is our Test. No one is really in that server but Aaron, I and a few others from time to time.
What’s the Eviction Policy set at?
Maybe but I haven’t seen that before with Epicor
I’ve seen it with MCV4 Web API on 2008R2 where Memory Cache was set to Aggressive, but not with 2012 or 2016.