This has been a good thread, even if it is a bit long, and it was quite
expedient for me, as I had a client just ask me about how to determine sortwork
sizes needed for a large database file they are processing. The records are
coming from inside a program via E15/35, so I can't use the standard SORT input.
After reading all these notes and going through the manual, I have built an
option override, but wanted to get your opinions on it. To all you experts out
there, are there any potential "gotchas" with using this?
//DFSPARM DD *
OPTION MSGDDN=SORTMSGS,MSGPRT=ALL,DYNALLOC=(SYSDA,9),DYNSPC=768,
HIPRMAX=OPTIMAL,MOSIZE=MAX,MOWRK=YES,DSPSIZE=MAX
/*
Also, I assume there is no way to force a Hipersort or a Dataspace sort? In one
of my tests this morning, I saw this:
ICE253I 0 RECORDS SORTED - PROCESSED: 456789779, EXPECTED: 456944716
ICE165I 0 TOTAL WORK DATA SET TRACKS ALLOCATED: 231120 , TRACKS USED: 0
ICE199I 0 MEMORY OBJECT USED AS MAIN STORAGE = 0M BYTES
ICE299I 0 MEMORY OBJECT USED AS WORK STORAGE = 17474M BYTES
ICE399I 0 MEMORY OBJECT USED IN ZSORT PATH = 0M BYTES
ICE180I 0 HIPERSPACE STORAGE USED = 0K BYTES
ICE188I 0 DATA SPACE STORAGE USED = 0K BYTES
Should I be concerned about the ICED180 and 188 messages showing they weren't
used?
Thanks for looking at this!
Don
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
