MCH4429 Automatic Storage Overflow on VALENCE6P job

[gwilburn]

I recently added a button to app. This calls a series of backend programs & service program procedures that interact with a web service. Multiple users are clicking on this throughout the day.
I have experienced two problems.
1. The Storage Overflow error - it references my SRVPGM and procedure

From program . . . . . . . . . :   Ex2C1DAutoStackOverflow
   Instruction  . . . . . . . . :     0001A0
 To program . . . . . . . . . . :   MKTR4
   To library . . . . . . . . . :     ECOM
   To module  . . . . . . . . . :     MKTR4
   To procedure . . . . . . . . :     UPDATEACCTNBR
   To statement . . . . . . . . :     62763
 Thread . . . . . . . . . . . . :   00000021
 Time sent  . . . . . . . . . . :   14:39:06.285428
 Message . . . . :   Automatic storage overflow.
 Cause . . . . . :   One of the automatic storage stacks X'00000002' for the
   thread has overflowed or a storage access beyond the maximum size of a
   teraspace automatic storage stack was attempted. Further program execution
   within the thread is not possible. Automatic storage stack values and their
   meanings follow:
     1 -- System stack in single level storage.
     2 -- User stack in single level storage.
     3 -- System stack in teraspace.
     4 -- User stack in teraspace.
 Technical description . . . . . . . . :   Attempt to reduce the automatic
   storage used by programs running in the thread.​

2. Record locks on the database tables updated by my process

I've never seen this before... it's like my program/procedures do not "end" - i.e. releasing the file locks and unloading the procedures from memory?

More info:
The same programs run just fine when submitted to a single-threaded job queue
I'm running in a named activation group

[robert.swanson]

Hi Greg, hard to say what's happening there, but since you're running your process in a named activation group you could try having a program higher in your stack (but in a different activation group) make a call to vvUtility_cleanUp to reclaim your activation group and clear the locks once it's all done. That could at least serve as a workaround until you can isolate the root cause.