|
Re: Memory linits for Unix IDL [message #10309 is a reply to message #10217] |
Fri, 07 November 1997 00:00  |
alpha
Messages: 49 Registered: September 1996
|
Member |
|
|
Bruce Macintosh <bmac@igpp.llnl.gov> writes:
> Stein Vidar Hagfors Haugan wrote:
>> John Varsik wrote:
>>>
>>> Is there a limit to how much virtual memory (other than the
>>> total amount of swap space) the Unix versions of IDL will use?
>>> If so, how do I set it?
>>
> I don't think IDL's use of memory is set just by available
> swap space.
right,
we have 1 GB over RAM and 2 GB of swap...
we worked with IDL using 1.5 GB with arrays at the same time..
slow but stable
Panther
--
Panther in the Jungle __..--''``\--....___ _..,_
-BELIEVE AND DECEIVE- _.-' .-/"; ` ``<._ ``-+'~=.
http://www.ang-physik _.-' _..--.'_ \ `(^) )
.uni-kiel.de/~hendrik ((..-' (< _ ;_..__ ; `'
|
|
|
Re: Memory linits for Unix IDL [message #10315 is a reply to message #10217] |
Fri, 07 November 1997 00:00  |
Achim Hein
Messages: 42 Registered: February 1996
|
Member |
|
|
Bruce Macintosh wrote:
>
> Stein Vidar Hagfors Haugan wrote:
>>
>> John Varsik wrote:
>>>
>>> Is there a limit to how much virtual memory (other than the
>>> total amount of swap space) the Unix versions of IDL will use?
>>> If so, how do I set it?
>>
>> Use
>>
>> unix> limit
>>
>> to see which limits are available. On e.g., Digital Unix there
>> is a limit for "vmemoryuse", which determines the amount of
>> virtual memory (i.e., swap space) that may be used.
>>
>> Stein Vidar
>
> I don't think IDL's use of memory is set just by available
> swap space. I've tried running on Sparcs with plenty of swap space
> (and vmemoryuse "unlimited" according to limit) but
> little physical memory (32 or 64MB), and I can't create more than
> a couple of 256x256x100 real arrays before running out of space.
> Otherwise-identical machines with 128MB or 256MB can create many more
> arrays. IDL seems to be aware in some fashion of how much physical
> memory is available, and unwilling to use more than that amount.
> (Note that, of course, with many users/processes it'll be forced
> to swap anyway...)
>
> Is this documented anywhere?
>
> Bruce Macintosh
> bmac@igpp.llnl.gov
There is a help file in IDL5 'Environment Variables Used by IDL'. There
you can read i.e. how to control the initial size of memory allocated to
hold IDL arrays and the amount of memory allocated when the array memory
block must me extended dynamically. BUT IT IS SHOWN FOR VMS - AND IT
DOESN'T WORK - AND I DON'T KNOW WHY sorry ?
Achim Hein
--
_______________________________________________
Dipl.-Ing. A. Hein
PB2 / ZESS - Uni-GH-Siegen
Paul-Bonatz Str. 9-11
57068 Siegen
Phone: 0271/740-3362
Fax: 0271/740-2336
Mail: Hein@nv.et-inf.uni-siegen.de
_______________________________________________
Please have a look at our Web-Sites:
http://www.nv.et-inf.uni-siegen.de/pb2/www_pb2
_______________________________________________
|
|
|
Re: Memory linits for Unix IDL [message #10324 is a reply to message #10217] |
Thu, 06 November 1997 00:00  |
Bruce Macintosh
Messages: 5 Registered: March 1997
|
Junior Member |
|
|
Stein Vidar Hagfors Haugan wrote:
>
> John Varsik wrote:
>>
>> Is there a limit to how much virtual memory (other than the
>> total amount of swap space) the Unix versions of IDL will use?
>> If so, how do I set it?
>
> Use
>
> unix> limit
>
> to see which limits are available. On e.g., Digital Unix there
> is a limit for "vmemoryuse", which determines the amount of
> virtual memory (i.e., swap space) that may be used.
>
> Stein Vidar
I don't think IDL's use of memory is set just by available
swap space. I've tried running on Sparcs with plenty of swap space
(and vmemoryuse "unlimited" according to limit) but
little physical memory (32 or 64MB), and I can't create more than
a couple of 256x256x100 real arrays before running out of space.
Otherwise-identical machines with 128MB or 256MB can create many more
arrays. IDL seems to be aware in some fashion of how much physical
memory is available, and unwilling to use more than that amount.
(Note that, of course, with many users/processes it'll be forced
to swap anyway...)
Is this documented anywhere?
Bruce Macintosh
bmac@igpp.llnl.gov
|
|
|