Originally posted by: ediX12HIPAA

I am confused with two of the suggestions between HANDLEIN and ECHOIN.

Beginning of this year, we had problem with a large file (1.5 GB), the code was using RUN function with "ECHOIN", the system was crashed.  We opened the ticket with IBM Support, we were suggested to use the function "HANDLEIN" because it can handle the large file.  There is a link explaining it:

https://www.ibm.com/support/knowledgecenter/en/SSVSD8_9.0.0/com.ibm.websphere.dtx.funcexp.doc/references/r_funcs_express_HANDLEIN.htm

Then on 8/5/2018, the IBM Newsletter had this article:

PI97740: CRASH OR DISK WRITE ERROR OR COULD NOT OPEN WORK FILES WHEN USING HANDLEIN() FUNCTION TO CALL MORE THAN 64K RUN MAPS

APAR status

• Closed as program error.

   

Error description

• When using the HANDLEIN() function to call more than 64K RUN
  map occurrences, one of the following problems may occur.
  
  1) A Crash
  
  2) Disk write error - return code 16
  
  3) Could not open work files - return code 36 (meaning - A
  scratch file for the HANDLEIN function could not be created.)
  
  
  The problem does not occur is the HANDLEIN() function is
  replaced with the ECHOIN() function.
  

https://www-01.ibm.com/support/docview.wss?uid=swg1PI97740&myns=swgother&mynp=OCSSVSD8&mync=E&cm_sp=swgother-_-OCSSVSD8-_-E

I am confused with these two issues and suggestions.  Please advise.

Best regards,

Huong

Originally posted by: PaulBrett

In general HANDLEIN() should be used in place of ECHOIN() because it does not create a copy of the data (merely a pointer to the original) and is therefore more conservative with memory usage.

The APAR shows that there was a small bug in the HANDLEIN() function, which could be avoided, if ECHOIN() was used instead.  The bug was fixed, and the ideal scenario would be to apply a patch to resolve the issue with HANDLEIN() and continue to use it.

Thank you

Paul

Follow me on Twitter

Originally posted by: ediX12HIPAA

Thanks Paul.