Hi Mark

I did wonder this but there is only 728 jobs in the system. The system has a tool called OSD running on it which clears down outq’s into an archive and then it’s backed up to tape. I don’t know if this is not clearing them down correctly?

Thanks

Darren

Hi,

I guess that’s not it then.

When did you last IPL? and how often has the machine been IPL’d before that?

Regards,

Martin Gilbert.

Martin

The last IPL was at the begining of this month, but I don’t think it’s something that has been done regularly on this system and could be even less than once a year.

Thanks

Darren

Hi,

Maybe it’s an idea to IPL a few times over the next few weeks. During an IPL the system runs a lot of different cleanup jobs, but some of these jobs need to run a few times (for example the first IPL marks things ready to remove, the following IPL really removes them).

It maybe an idea to schedule IPL’s when the system has some idle time (in the weekend or on a friday evening maybe). IPL’s help keep the system running smoothly and cleanly.

It might also be worth running another RCLSTG. A RCLSPLSTG is also an idea.

Regards,

Martin Gilbert.

Hi,

Just another thought – try running the CLEANUP jobs :-

GO CLEANUP

Maybe it’ll help.

Regards,

Martin Gilbert.

Try WRKJRNRCV *all and look for an excessive number of journal receivers or very large receivers that haven’t been cycled. Running Cleanup with the right parms

Change Cleanup Options CSYS
04/21/09 12:54:56
Type choices below, then press Enter.

Allow automatic cleanup . . . . . . . . . . . . . Y Y=Yes, N=No

Time cleanup starts each day . . . . . . . . . . 22:00:00 00:00:00-
23:59:59,
*SCDPWROFF,
*NONE

Number of days to keep:
User messages . . . . . . . . . . . . . . . . . 7 1-366, *KEEP
System and workstation messages . . . . . . . . 4 1-366, *KEEP
Critical system messages . . . . . . . . . . . *KEEP 1-366, *KEEP
Job logs and other system output . . . . . . . 7 1-366, *KEEP
System journals and system logs . . . . . . . . 30 1-366, *KEEP

should get rid of system journals but not user journals if there are any. Also check for savfs. Maybe do a dspobjd *allusr/*all *files to an outfile and query if for savfs. You may have old savfs that haven’t been cleared.

Thanks everyone for your reply.

We have tried the cleanup but it had no effect and it is already set to automatic as suggested by DanD. We have no big or old journal receivers on the system. I have also run the RCLSPLSTG with no efect.

We have scheduled a RCLSTG *DBXREF for this weekend and an IPL to see if that helps and then a full RCLSTG the following weekend with another IPL to see if that helps.

If that doesn’t work then it’s time to think up a plan B. Any suggestions welcome.

Do you have a bunch of objects like these on your system? We regularly run a job to get rid of the older of these objects. They can add up pretty quickly. The output illustrated below is from PRTDSKINF.

QPMJ02T350 QPFRDATA *DTAQ QSECOFR .01 17850.4 04/12/09 03/16/09
QPMJ04T937 QPFRDATA *DTAQ QSECOFR .01 17850.4 04/12/09 03/16/09
QPMJ05T003 QPFRDATA *DTAQ QSECOFR .01 17850.4 04/12/09 03/16/09
QPMJ07T936 QPFRDATA *DTAQ QSECOFR .01 17850.4 03/15/09 02/23/09
QPMJ08T133 QPFRDATA *DTAQ QSECOFR .01 17850.4 03/15/09 02/23/09
QPMJ50T061 QPFRDATA *DTAQ QSECOFR .01 17850.4 03/15/09 03/02/09

Any chance there are a bunch of PTF files on your system? This text is from the command itself:

“The Delete Program Temporary Fix (DLTPTF) command is used to delete program temporary fix (PTF) save files, their associated cover letters, and the records of PTFs that have been ordered.”

Whatever is taking up the room, it’s not a visable object. We have 173Gb of disk and 84% (145Gb) being used. When I run a PRTDSKINF *LIB, the total at the bottom of the report of all the librarys added together is about 40Gb. So the rest of the 100Gb is in black hole somewhere called “internal objects – manage jobs”.

Are you running a RTVDSKINF before the PRTDSKINF? Is the RTVDSKINF being run when the system is restricted or very very low activity? What is the date the data was collected on the first page of the PRTDSKINF report?

I think I might have found the answer, I always thought the bit at the bottom of page 3 of the Disk Info report showed how much disk space you would get back from a RCLSTG but A colleague of mine was looking on the IBM website and found the following “Note: The Managed Jobs field on A disk tasks report will show the amount of space a RCLSTG will return” see link below.

http://www-01.ibm.com/support/docview.wss?rs=0&dc=DB520&dc=D900&dc=D800&dc=DA900&dc=DA800&q1=Clear+Down+Managed+Jobs+on+AS400+AND+AS400KBXXYYZZRCH&uid=nas18a0e2b691bd0fb6b86256705007cb354&loc=en_US&cs=UTF-8&lang=all

I had a problem much like this. We IPL every month-end. Suddenly, in the middle of the month we are getting warnings about over 90% of DASD ( we normally run about 40%). Deleted a fewlibraries, but it didn’t go down much. The disk reports only added up to about the 40% value. So … 2 weeks into the new month, we did a unscheduled IPL. Disk went right back to 40%, and 2 months later we haven’t seen it again. Never did figure out exactly what the problem was.

Hi,

That’s pretty scary that you can suddenly “lose” 50% of your DASD for no apparent reason. Most of our non-production systems are normally around 80% full, so I guess if we get the same issue they’re going to go down?

Luckily our production systems also run around the 40 % full, so we should survive…maybe…

Regards,

Martin Gilbert.