Re: QDFTSVR job description question. -- MIDRANGE-L

Printouts are fun but for JobD's you could collect them to a "history" file.

something like:
create table your_lib.jobd_hist as(
select current_server as sysname, i.*,current_timestamp as current_time from qsys2.job_description_info i
where job_description like 'Q%') with data;

insert into your_lib.jobd_hist
(select current_server as sysname, i.*,current_timestamp as current_time from qsys2.job_description_info i
where job_description like 'Q%')

just for fun...

Bryan

Michael Mayer wrote on 2/9/2023 4:34 PM:

It’s times like these when someone says …

“That tells me that at some point the job description had been changed from
the shipped values. I don't recall doing it and don't know who would”.

… That I feel good knowing I’ve run a PRTSYSINF, a SYSSNAP from the MG TOOLKIT and
I have a custom PRTJOBD CLLE that cranks out a spooled file of every jobd on our
systems. If anyone wants the code for that, send me an email.

I use ACS “Printer Output” to pull the spooled files to our LAN for safekeeping and
reference.

For the SYSSNAP, that output usually goes to IFS ‘/tmp’. I use the ACS IFS feature to
pull the info down to our LAN.

I do this every 6 months or after PTF’s and/or upgrades. or any time system info
changes.

Respectfully,
Michael Mayer
IBM i on Power System Admin
Cell: 518.641.8906
Office and On Call: 731-676-4318
michael.mayer@xxxxxxxxxxxxx<mailto:michael.mayer@xxxxxxxxxxxxx>
https://www.ermco-eci.com<https://www.ermco-eci.com/>
IBM i Blog: https://ibmireference.blogspot.com<https://ibmireference.blogspot.com/>

[cid:image001.gif@01D93CA3.66EE7660]

From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> On Behalf Of K Crawford
Sent: Thursday, February 9, 2023 4:07 PM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxxxxxxxx>
Subject: Re: QDFTSVR job description question.

Dawn, thank you for the detailed explanation.
That tells me that at some point the job description had been changed from
the shipped values. I don't recall doing it and don't know who would. In
any case I change it back to the shipped values. That sounds like what I
wanted anyways. I had to keep an eye on the number of spool files and
clean them on occasion. Hopefully I will not need to now.
Once again thank you. This forum has been a wealth of information for me.
Thanks to all of you.
Thanks Dawn
Kerwin

On Thu, Feb 9, 2023 at 3:42 PM Dawn May <dawnmayican@xxxxxxxxxxx<mailto:dawnmayican@xxxxxxxxxxx>> wrote:

One correction - the client that hit the 2.6 million spooled files was
using LOG(4 0 *SECLVL); my original post did not state that correctly.

And as long as I’m writing this update, you can track the number of
spooled files in the system ASP using the System Health Services.
The maximum number of spooled files in the system ASP is also one of the
alertable limits and an SQL7062 message will be sent to QSYSOPR if you are
getting close to the maximum.

System Health Services in the Knowledge Center has more information.
https://www.ibm.com/docs/en/i/7.5?topic=services-system-health<https://www.ibm.com/docs/en/i/7.5?topic=services-system-health>

I have written several iCan blog posts on system health services and
system limits.
If you go to the iCan Blog of Blogs and use find to search the page for
“limits” you can easily find these posts.
https://dawnmayi.com/i-can-blog-of-blogs/<https://dawnmayi.com/i-can-blog-of-blogs>

Dawn

On Feb 9, 2023, at 3:30 PM, Dawn May <dawnmayican@xxxxxxxxxxx<mailto:dawnmayican@xxxxxxxxxxx>> wrote:

As shipped by IBM, the QDFTSVR job description has LOG(4 00 *NOLIST),

LOGCLPGM(*NO), and ENDSEV(30), and in general, the QDFTSVR job description
should not be changed from these settings. The QDFTSVR job description
settings were carefully selected by IBM for optimal system performance and
health characteristics for prestart jobs. These settings allow for all
messages to be logged to the job log while it is running; if the job ends
with a severity of 30 or greater, the job log will be created. If the job
ends normally, then no jog log is created.

It sounds like changing QDFTSVR back to the shipped settings is what you

want to do.

When logging is turned on for a prestart job, a job log will be created

every time a user profile swap occurs in one of those jobs.

I want to emphasize that point - a job log will be created every time a

user profile swap occurs.

Since the database server prestart jobs can be reused up to 200 times,

and database connections can be frequent, this can result in a very large
number of job logs. In addition, the QDFTSVR job description is specified
in several prestart job entries, so change the job description impacts all
the prestart jobs that use that job description. I worked with a client
that started hitting the maximum number of spooled files in the system ASP
- which is 2.6 million - because of LOG(4 00 *NOLIST) on QDFTSVR. You do
not want that to happen to you!

If you need logging enabled in order to do troubleshooting, you should

not change the QDFTSVR job description. Rather, you should create a
duplicate of it and change the settings you need to modify. Then, change
the prestart job entry of ONLY the jobs you are troubleshooting to use the
new job description. You can change the prestart job entry without ending
the prestart jobs; if that is done, active prestart jobs at the time of the
change continue to run with the job description used when the job was
started. The new job description gets picked up as new prestart jobs are
started over time. To have the job description change take effect
immediately, end and restart the prestart jobs (END/STRPJ commands).

When logging is turned on for prestart jobs, you need to carefully watch

the job logs on your system so ensure you don’t start to get near the limit
of the maximum number of spooled files.

When you are done troubleshooting, you should change the prestart job

entry back to use the QDFTSVR job description.

Dawn

On Feb 8, 2023, at 3:56 PM, K Crawford <kscx3ksc@xxxxxxxxx<mailto:kscx3ksc@xxxxxxxxx>> wrote:

We have many QZDASOINIT and QRWTSRVR jobs by user QUSER using job
description QGPL/QDFTSVR. These create many many QPJOBLOG spool files

each

day. Vary few of them have any true errors in them. I only want spool
files if something happens.
Would I be better off changing that job description or copying it and
changing the jobs/users to use the new one? Will be a lot easier to

change

it. Being in QGPL I think it would be safe during our next upgrade.

I want a log if it fails but nothing if no errors.

It is currently set up like this.
Message logging:
Level . . . . . . . . . . . . . . . . . . . . : 4
Severity . . . . . . . . . . . . . . . . . . . : 0
Text . . . . . . . . . . . . . . . . . . . . . : *SECLVL
Log CL program commands . . . . . . . . . . . . : *YES

Thinking of changing to this. Main reason I do get some level 40 saying
library already in library list.
Message logging:
Level . . . . . . . . . . . . . . . . . . . . : 4
Severity . . . . . . . . . . . . . . . . . . . : 41
Text . . . . . . . . . . . . . . . . . . . . . : *SECLVL
Log CL program commands . . . . . . . . . . . . : *YES

Thanks
--
KCrawford
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx<mailto:MIDRANGE-L@xxxxxxxxxxxxxxxxxx>
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l<https://lists.midrange.com/mailman/listinfo/midrange-l>
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx<mailto:MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx>
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l<https://archive.midrange.com/midrange-l>.

Please contact support@xxxxxxxxxxxxxxxxxxxx<mailto:support@xxxxxxxxxxxxxxxxxxxx> for any subscription

related questions.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx<mailto:MIDRANGE-L@xxxxxxxxxxxxxxxxxx>
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l<https://lists.midrange.com/mailman/listinfo/midrange-l>
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx<mailto:MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx>
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l<https://archive.midrange.com/midrange-l>.

Please contact support@xxxxxxxxxxxxxxxxxxxx<mailto:support@xxxxxxxxxxxxxxxxxxxx> for any subscription related
questions.

--
KCrawford
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx<mailto:MIDRANGE-L@xxxxxxxxxxxxxxxxxx>
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l<https://lists.midrange.com/mailman/listinfo/midrange-l>
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx<mailto:MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx>
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l<https://archive.midrange.com/midrange-l>.

Please contact support@xxxxxxxxxxxxxxxxxxxx<mailto:support@xxxxxxxxxxxxxxxxxxxx> for any subscription related questions.

The information in this email may be confidential. It is intended only for the person(s) named above. If you are not the intended recipient, please notify the sender and do not review, distribute, or duplicate this email.