Re: Question on QTEMP and CGI Jobs

Thanks for the reply.

Since the CGI request/response cycle is synchronous and performing a single CGI transaction why would steering away from QTEMP be bad ? I'm not following your time slicing theory.

Ultimately I may suggest that the developers place the additional session data into the session database we already have, but your response seems contradictory on whether QTEMP is unique or not.

You seem to advise against it, yet isn't QTEMP a hallmark of IBMi jobs and unique across jobs as we've been discussing ?

Just curious :-)

Regards,
Richard Schoen
Director of Document Management
e. richard.schoen@xxxxxxxxxxxxxxx
p. 952.486.6802
w. helpsystems.com

---------------------------------
Subject: Re: Question on QTEMP and CGI Jobs
From: Nathan Andelin <nandelin@xxxxxxxxx>
Date: Sat, 4 Mar 2017 08:52:44 -0700

I don't see any way a qtemp object could get clobbered by another job. As
others have indicated the request-response cycle is synchronous. But since
a single CGI job can have an unlimited number of programs active in the
same job, it seems theoretically possible that one of them could clobber a
qtemp object due to CPU time slicing. Perhaps unlikely, but I'd still steer
away from using qtemp in CGI jobs.