Question

MRP run time doubled, throws error "The socket connection was aborted."

  • 11 August 2022
  • 2 replies
  • 112 views

  • Anonymous
  • 0 replies

10.2.700.21
Has anyone experienced the below issue before? We have been getting the attached error concerning a socket timeout without warning, after going several days with MRP running successfully. Run time is approx. 4 hrs. normally for a regenerative run, however when the error occurs run time at least doubles to 8 hrs. or longer, and MRP continues to run in the background while The process reports the error in System Monitor, then will change to “Complete” in System Monitor when it finishes. Any help or insight on this would be greatly appreciated!
 


2 replies

Userlevel 2

This thread may be related:

 

We have experienced the same issue.  The reason MRP takes longer when you get that is because the number of processors have decreased.  Setting the number of processors and schedulers is a bit of a science based on your setup. 

In our environment, we run 5 processors, 3 schedulers.  The processors handle the parts at the given level and rotate the level until finished, then move on to the next level.  If we lost 1 processor, that’s 20% of our bandwidth.

I noticed the abandon was right before midnight.  If the time is consistent then I would check to make sure something else isn’t going on with the system at that time (e.g. a VM snapshot used to cause issues with task agent back in the day).  

Jenn

Reply