QRWTSRVR job starting?

[tupurz]

We are currently running Net.Data on a linux box to fill web page
queries of AS/400 data. Every couple of days, I get a job going by the
name of QRWTSRVR that is attemtping to index the table that I am doing
the query on. This job start sucking up CPU cycles; it can run for
days and it continues to grow eating up more cycles...if it doesn't
complete, another QRWTSRVR may get started, and maybe, if we're lucky,
another one might get going.

I can only find two posts on the net that reference this guy and
neither is a big help.

If anyone can help it would be appreciated.

Shawn

Sent via Deja.com http://www.deja.com/
Before you buy.

[Karl Hanson]

The QRWTSRVR jobs handle remote SQL server-side requests using a
protocol called DRDA (over TCP/IP). In this case the client side
(Applicatin Requester in DRDA lingo) is probably the Linux box. When
you see one of these long-running jobs, you might try getting a job log
with more detail to see what it is doing. If you do this:

===> chgjob nnnnn/QUSER/QRWTSRVR log(4 00 *seclvl)
===> strsrvjob nnnnn/QUSER/QRWTSRVR
===> strdbg updprod(*yes)

... additional informational messages will be generated in the job log
as to what SQL operations are being done.
Note that these are prestart jobs that normally get recycled, so to get
a QPJOBLOG spool file with the info messages, you may need to:
A) Create one manually before the job is recycled:
===> dspjoblog nnnnn/QUSER/QRWTSRVR output(*print)
-or-
B) Temporarily change the subsystem prestart job entry so the job is not
recycled:
===> CHGPJE SBSD(QSYS/QSYSWRK) PGM(QSYS/QRWTSRVR) MAXUSE(1)

However if you do change MAXUSE to 1 (per B above) remember to reset it
to the original value (normally 200); otherwise performance may be
degraded in the long term.

--

Karl Hanson