On Wed, Sep 30, 1998 at 11:26:13PM -0700, [EMAIL PROTECTED] wrote:
> On  1 Oct, Robert G. Brown wrote:
> > Hmm, never heard of swap not working before.  Are you sure you are
> > invoking it correctly?  Try the script below.  Note that with multiple
> > SCSI disks, swap will be a bit more efficient if the swapfiles are on
> > different disks instead of all in one partition.  Using swapfiles
> > instead of swap partitions conserves partitions -- since you need at
> > least four swap spaces to swap a full 512 MB (linux has a 128 MB limit
> > per) you would need to break up a single disk into 5+ partitions,
> > which is a pain.
> 
> Is using swap files any faster or slower than using swap partitions?
>   -M@
> 

 Hello,

IIRC using swap _files_ is discouraged on production systems,
as there is always a (very small) risk of deadlock and there is a performance
penalty.

If you need fast swap, use swap partitions on diferent disks,
and specify the same priority.

 Ciao
     Dietmar

quoting man 2 swapon:
PRIORITY
       Each swap area has a priority, either high  or  low.   The
       default  priority  is low.  Within the low-priority areas,
       newer areas are even lower priority than older areas.

       All  priorities  set  with  swapflags  are  high-priority,
       higher than default.  They may have any non-negative value
       chosen by the caller.  Higher numbers mean  higher  prior�
       ity.

       Swap  pages  are  allocated  from areas in priority order,
       highest priority first.  For areas with different  priori�
       ties,  a  higher-priority area is exhausted before using a
       lower-priority area.  If two or more areas have  the  same
       priority,  and it is the highest priority available, pages
       are allocated on a round-robin basis between them.

       As of Linux 1.3.6, the kernel usually follows these rules,
       but there are exceptions.



-- 
Reporter (to Mahatma Gandhi): Mr Gandhi, what do you think of Western
Civilization?  Gandhi: I think it would be a good idea.
Dietmar Goldbeck, E-Mail: [EMAIL PROTECTED]; phone +49-5241-80-7646

Reply via email to