[Info-vax] Process memory settings

[Syltrem]

"Joseph Huber" <joseph.huber at NOSPAM.web.de> wrote in message 
news:h61a24$l0$1 at gwdu112.gwdg.de...
> Syltrem wrote:
>
>>
>> "Syltrem" <syltremzulu at videotron.ca> wrote in message
>> news:AwUgm.549222$Tp1.144878 at en-nntp-01.dc1.easynews.com...
>>> Also if that helps, no matter what parameter I change (PGFLQUO, WSEXTENT
>>> etc) the SHOW PROC/QUO will always show this after the compiler exits :
>>>
>>> Peak working set size:     645568
>>> Peak virtual size:         845664
>>>
>>>
>>> And I cannot increase the max authorized WS so that "Peak virtual size"
>>> can go higher. Why can't I increase it ? Where is this value derived 
>>> from
>>> ?
>>
>> I tried a few more things:
>>
>> COBOL/NOOBJECT
>> This is not very helpful to me, but it does complete successfully so 
>> tells
>> the the source code is ok and compiler optimization can be done.
>>
>> COBOL/NOPTIMIZE also completes successfully. Better.
>>
>> COBOL/OPTIMIZE=LEVEL=1 fails.
>>
>> Though I now have something usable, if not optimal, I'd still want to
>> understand how to increase the working set, if someone knows.
>>
>
> As others have told already, the working set size is NOT your problem.
>
> But to answer the question: it is the minimum of WSEXTENT (per user in 
> UAF),
> and the system parameter WSMAX.
>
> --
> Joseph Huber, http://www.huber-joseph.de

Right. I looked at WSMAX yesterday but was not entirely sure. I would need 
to reboot to increase it.

I lowered signifiicantly the PQL_MWSEXTENT and that of the user in UAF
It induced a lot of page faults of course, and the Peak Virtual Size still 
grew to 847968, which is the same value I get to with higher settings of 
WSEXTENT.
It always fails when it needs more memory than this, whether from the real 
memory of page file.

WSMAX currently is set to 872448 which is close to this maximum Virtual Size 
I get. It looks like this is my upper boundary.
PQL_MWSEXTENT normally is equal to WSMAX but I lowered it for this test.

I think I'm going to install the cobol compiler on a test box that I can 
reboot, and play with WSMAX.
It looks like it has an incidence not only on the working set, but on the 
total memory used by the process.


Syltrem