Re: system is loosing disk space at extraordinary rate

[Bob Larkin <blarkin@xxxxxx>]

It is basic, but check for journal receivers that aren't being cleaned up, 
WRKPRB to see if you
have too many problems, check for QHST logs. as wel as VLOGs and PAL entries 
(in STRSST).  These
shouldn't affect temp space, but worth a check.
Bob

Philipp Rusch wrote:

> Memory leak in V4R5 ?
>
> A short summary of what we already investigated:
>
> - we controlled spool files / outqueues: there are few files (some hundred, 
>not unnormal amount
> for a system of this size)
> - we are not using Save While Active (SWA), neither in our own jobs, nor on
> the commercial packages we use.
> - we don't do SQL (this was the one with the big index files)
> - we dont't do TCPIP over SNA controllers, we only have one 5494 doing native 
>SNA
> - we dont't see high I/O activity on the system
> - we don't see jobs "abusing" the system with high percentages of CPU capacity
> - we are running a scheduled RTVDSKINF in the morning (6.00) and in the 
>evening (19.00)
> to be able to compare %temporary used against the values of the snapshots 
>before
> Results so far:
> - the nightjobs don't "eat" disk, seems to be no backup issue, then
> - even if only some users (under ten) are using the system (its easter 
>holidays in germany)
> it eats memory at a significant rate (200 MB during a day)
> - we have intensive use of IFS-directories for temporary storage of bitmap 
>files that are
> being merged into print jobs on the PC that triggers this job, could this be 
>the culprit ?
>
> Thanks everybody for your tips so far,
> Philipp Rusch
>
> Philipp Rusch schrieb:
>
> > Hello all,
> > this is a strange one:
> > I did a release upgrade on a V4R4 machine (a 720) to V4R5 in november last 
>year.
> > We were at about 50% used disk space at that time. PTF package C0198450 was
> > installed at that date as well, everything went fine.
> > Since we did not do any further IPLs since then, we someday noticed that 
>something
> > is eating up disk space, until we reached 92 % last week.
> > Searching for the culprit was at no avail, we ended every user jobs, ended 
>every non-system
> > job on the system, it eats space on and on. A short calculation led us to 
>round about
> > 300 MB (!) temporary space which is used up every day since november.
> > WRKSYSSTS showed 29 GB of temp used space out of 77 GB total, this is 38 % 
>of the
> > whole system !
> > When we scheduled an IPL to get rid of the temporary used space, I did 
>endsbs *all *immed
> > at first hand and suddenly all our used disk space was free again ! You 
>could watch it coming
> > back available when pressing F5 in wrkssysts display. I never saw this 
>behaviour on an AS/400
> > before. IBM support was nearly as helpless as we, we ordered some PTFs that 
>might help ...
> >
> > Any hints anybody ?
> >
> > Have a nice easter weekend,
> > regards from germany, Philipp Rusch
> >