Mike Cunningham,
A couple years ago our job tables got large and WRKSBMJOB command ran
very slowly.
Compressing the Job Tables really helped.
Our CHGIPLA is now set to:
Compress job tables . . . . . . CPRJOBTBL *NORMAL
IBM Tech Support comments are shown below.
eric
+++++ one time ++++++++++++++
Your Job Tables are quite fragmented. This is likely the cause of the
poor performance of any command that must search the job tables to
return results.
You can see that you have 5 tables 4 full and 1 partial, totalling 67100
job structures. But of those 67100 structures, only 6131 are in use.
I would recommend compressing the job tables. You can do this at IPL
time. Before the IPL, use the CHGIPLA and set Compress Job Tables to
*NORMAL. Then after a normal IPL make sure it gets set back to *NONE.
One WARNING though - job table compression could add significant time to
your IPL. We've seen hours but in your case I would guess around 15-45
mins but that is only a guess.
If compressing the job tables does not solve this, we will likely need
to install a performance tool and collect some data. But before we do
that let's see how these commands work with compressed job tables.
++++++ another time +++++++++++++
We typically don't recommend compressing the job tables unless the
customer is experiencing storage or performance problems due to the
quantity of job structures. Any performance problems would be seen with
commands such as WRKSBMJOB or WRKUSRJOB, which need to run the job
tables in order to return information. Other than using up some
storage, these additional job structures are not causing any problems on
the system.
Compressing the job tables will impact the IPL performance. The
additional IPL steps can easily add 1 or more hours to the IPL time. We
don't have any way to estimate the performance impact. We would just
recommend waiting to compress the job tables until some time when the
IPL can run longer than usual.
There are two steps to compressing the job tables.
First, the in-use entries are moved to the front of the tables. This
is the step which takes the large amount of time. On your system
the in-use entries are already in table 1, so this step may run
quickly.
The second step is to delete the extra available entries down to the
QTOTJOB value. This would delete tables 2 - 5 and truncate table 1
down to 9000 entries.
After compressing the job tables, you will want to set the IPL attribute
back to *NONE. There is no need to run this step every IPL.
16.5 IPL Tips
Although IPL duration is highly dependent on hardware and software
configuration, there are tasks that
can be performed to reduce the amount of time required for the
system to perform an IPL. The following
is a partial list of recommendations for IPL performance:
! Remove unnecessary spool files. Use the Display Job Tables
(DSPJOBTBL) command to monitor
the size of the job table(s) on the system. Change IPL Attributes
(CHGIPLA) command can be used
to compress job tables if there is a large number of available job
table entries. The IPL to compress
the tables will be a long one, so try to plan it along with a
normal maintenance IPL where you have
the time to wait for the table to compress.
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact
[javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.
midrange.com
