Re: shared arena size

New Message Reply Date view Thread view Subject view Author view

Angus Dorbie (dorbie++at++sgi.com)
Sat, 07 Aug 1999 02:54:53 -0700


Good point, the shared arena size passed as an environment variable but
it's just an indicator to performer to request those system resources.
The rlimits would have to be adjusted so the kernel permitted the
allocation of those resources or the request will not be honoured.

Cheers,Angus.

Robert Wolfe wrote:
>
> Does it serve any purpose to use systune to adjust
> system resource sizes such as rlimit_rss_max or
> rlimit_rss_cur in order to be able to increase
> PFSHAREDSIZE ?
>
> Robert
>
> Angus Dorbie wrote:
> >
> > There is no easy answer, and a couple of issues you must consider.
> >
> > The arena must be contiguous in virtual address space and the loading of
> > various dso's can adversely affect the arena growth.
> >
> > Making the arena large to begin with is helpfull but remember it grows
> > by doubling so you may see counterintuitive results if you don't bear
> > this in mind.
> >
> > You should check the archives for detailed information on moving DSO's
> > to support large databases. It seems like those cases are a pretty good
> > match with yours.
> >
> > Cheers,Angus.
> >
> > Klaus Otto wrote:
> > >
> > > Hi,
> > >
> > > I'm trying to load a MultiGen database of about 500MB of flt-files
> > > into perfly on a 512 MB Memory Onyx2 IR, Irix 6.52.
> > >
> > > After a while, perfly terminates telling me:
> > > unable to allocate x bytes from arena ...
> > >
> > > I increased PFSHAREDSIZE and swap space on disk with some strange
> > > results given in the list below:
> > >
> > > HDisk swap PFSHAREDSIZE Terminated at
> > > 2.5 gb 1950000000 962000 kB
> > > 2.5 gb 1999999000 976000 kB
> > > 2.5 gb 2500000000 610351 kB
> > > 2.5 gb 1000000000 976000 kB
> > > 1.5 gb 1999999000 976000 kB
> > > 3.5 gb 2500000000 870993 kB
> > > 3.5 gb 2200000000 934246 kB
> > >
> > > It sometimes looks like decreasing the PFSHAREDSIZE gives more space.
> > >
> > > I know from the achives that this has been topic several times,
> > > but none of those cases seem to match this one.
> > >
> > > Thank you in advance,
> > >
> > > Klaus
> > > ------------------------------------------------------
> > > | Klaus Otto | |
> > > | CAE Elektronik GmbH | Email: kotto++at++cae-gmbh.de |
> > > | Abt. Systemtechnik | |
> > > | Steinfurt 11 | Tel. : +49-2402-106-456 |
> > > | D-52222 Stolberg | Fax : +49-2402-106-270 |
> > > | Germany | Web : http://www.cae.de |
> > > ------------------------------------------------------
> > >
> > > -----------------------------------------------------------------------
> > > List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> > > Submissions: info-performer++at++sgi.com
> > > Admin. requests: info-performer-request++at++sgi.com
> >
> > --
> > Never express yourself more clearly than you think.
> > Neils Bohr
> > Read this:
> > http://www.vcnet.com/bms/features/3d.html
> >
> > For advanced 3D graphics Performer + OpenGL based examples and tutors:
> > http://www.dorbie.com/
> > -----------------------------------------------------------------------
> > List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> > Submissions: info-performer++at++sgi.com
> > Admin. requests: info-performer-request++at++sgi.com
>
> --
> Robert Wolfe, consultant
> Simulation and Training
> Defence and Civil Institute of Environmental Medicine
> 1133 Sheppard Ave., W.,
> Toronto, Ontario
> CANADA
> M3M 3B9
> (416) 635-2065
> robert.wolfe++at++dciem.dnd.ca
> -----------------------------------------------------------------------
> List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com

-- 
Never express yourself more clearly than you think.
                                       Neils Bohr 

For advanced 3D graphics Performer + OpenGL based examples and tutors: http://www.dorbie.com/


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Sat Aug 07 1999 - 02:55:03 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.