• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1267
  • Last Modified:

Running out of disk space

I am monitoring a couple of iSeries systems and one of them is running out of disk space. I am not an admin though, but asked by him to post here!

One machine runs at 92% ASP and it is hard to find out what the heck is filling up the machine. We have got like 400k SPOOL files, which is too much and will be deleted soon, but that does not seem to be the problem. An IPL is scheduled for this weekend anyways.

System is still running on V5R3.

Is there a way to see this in iSeries Navigator or some command that shows the libraries and files / subfiles with their sizes?
0
Ferrosti
Asked:
Ferrosti
  • 3
  • 3
  • 2
  • +1
3 Solutions
 
Gary PattersonVP Technology / Senior Consultant Commented:
0
 
Jon SnydermanCommented:
To really get down to what is taking the space, you can use the RTVDSKINF and PRTDSKINF commands.  The latter cant be run interactively.  You need to submit it.

To submit the analysis: SBMJOB CMD(RTVDSKINF) JOB(RTVDSKINF)
To check its status: WRKSBMJOB and look got job: RTVDSKINF.
When the job is in OUTQ, run PRTDSKINF and prompt it with F4.  

This will break down all the disk space on the system and help you find the bullk of the matter.

~ Jon
0
 
FerrostiAuthor Commented:
As far as I know our admin is going to do the IPL as well as the Reclaim Disk Storage. As far as I understood it (I am more familiar with Unix), this will need to be done in restricted mode right after IPL.

'Disk Tasks' is said to need at least 3% of disk space (link 1). At this point this will certainly blow up the machine and this one is the most important. Before the weekend we were down to 89% and today it got up to almost 93%.

We are archiving spool files to an external DMS system like invoices, orders, etc. Could it be possible that deleted spool files space is not reclaimed automatically by the system?
0
VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

 
LewisPowerCommented:
Ferrosti,

What's interesting with jsnyderman command is once the RTVDSKINF has finisehd to run (it could take several hours), you could print a first report (PRTDSKINF RPTTYPE(*LIB) by example), then refine your searches (PRTDSKINF RPTTYPE(*LIB) LIB(XYZLIB) OBJ(*ALL) by example). It can be sorted by objet size, etc. This command is use more to decrease your datas.

What's interesting with Gary link, it's to decrease space from the OS environment. It will help you delete performace collection objects, reclaim storage, etc.

Both ways are very interesting !

Lewis
0
 
Jon SnydermanCommented:
Running the RTVDSKINF has a line that will show exactly how much space will be regained by running a RCLSTG.    It will also show you specifically how much the OS is taking for spooled objects, history logs, etc.  

I would suggest running that before you hang your hat on an IPL and a RCLSTG.  

-Jon
0
 
Gary PattersonVP Technology / Senior Consultant Commented:
Clean up your spooled files, then run a RCLSPLSTG (0).  Spooled file space is not automatically released by the system.  The OS assumes that if you had 400K spooled files recently, you'll need space for them again.

Any time you make a drastic reduction in spooled files, or change you spool archival strategy, or have a runaway job that creates hundreds of thousands of unwanted spooled files, you should do a RCLSPLSTG.

http://www-01.ibm.com/support/docview.wss?uid=nas1edf0ff7a4ec033c186256fc1007d2d28

If you just follow the system cleanup instructions in my first reply, it covers most of what you need to know.  If you still have space problems, you might want to engage the services of an AS/400 expert that can do a more detailed analysis for you.

- Gary Patterson
0
 
FerrostiAuthor Commented:
Thanks a lot for all your help.
Our Admin does not want to do all this before the upcoming weekend and the scheduled IPL. As far as I understood him this will be taken into consideration, but due the fact for the machine beeing extremely slow he does not want to do anything on it reducing performance during working hours.
After working hours and start of nightly batch jobs there is not enough time to achieve all this.
0
 
Gary PattersonVP Technology / Senior Consultant Commented:
The real answer to your question is in the two top links that I provided in my first response.

Tell your admin that there are lots of of small, fast, easy things he can do to cleanup during normal hours without waiting for the weekend or an IPL.  
Again, refer to the first link up top for details on these:

  • QHST cleanup
  • CHGCLNUP
  • Delete joblogs and debug spooled files
  • Backup and delete journal receivers.
  • Clean up unneeded DLO's, and run RCLDLO (I'd submit the RCLDLO to batch, though)
  • Clean up any unneeded old performance data
  • Spooled file cleanup and RCLSPLSTG can go at any time.
  • SQL log cleanup
  • Problem log cleanup
  • Index advisor file cleanup
  • Stop SMTP tracing if running and cleanup traces.
These items are all non-performance intensive, and many can be accomplished in just a few minutes.  Silly not to do them now if you're having space and performance issues.

- Gary Patterson


0
 
Jon SnydermanCommented:
I second that.  Gary is absolutely right.  I would encourage you to take action asap.  An iSeries running out of space is very bad.  Also, my post about running the RTVDSKINF can be submitted at low priority to run in the background without impacting your users.  

~ Jon
0

Featured Post

Free Tool: SSL Checker

Scans your site and returns information about your SSL implementation and certificate. Helpful for debugging and validating your SSL configuration.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

  • 3
  • 3
  • 2
  • +1
Tackle projects and never again get stuck behind a technical roadblock.
Join Now