[OpenAFS-devel] fileserver profiling

Troy Benjegerdes hozer@hozed.org
Sun, 20 Mar 2005 09:18:37 -0600


On Wed, Mar 16, 2005 at 04:59:47PM -0500, Jeffrey Hutzelman wrote:
> 
> 
> On Monday, March 14, 2005 07:28:54 AM -0500 chas williams - CONTRACTOR 
> <chas@cmf.nrl.navy.mil> wrote:
> 
> >In message <20050314040142.GN9768@kalmia.hozed.org>,Troy Benjegerdes
> >writes:
> >>I just switched to a 100mb cache, (previously it was 1GB), and ran afsd
> >>with "afsd -memcache -chunksize 16", and now I'm getting a better, but
> >>not optimal throughput of around 25MB/sec.
> >>
> >>(Out of curiosity, what cache sizes do people generally run at?)
> >
> >generally speaking memcache is going to be faster
> 
> ... but less storage-efficient
> 
> a 1GB disk-cache will store considerably more than a 1GB memcache, unless 
> the chunksize is no larger than 1K.

So if I have a 1GB memcache, I assume this means the memory is
completely unavailable for anything else?

The linux buffer/page cache works quite nicely with NFS, and on a
machine with several gigs of memory, if all you are doing is compiling,
everything gets cached.

What I'd really like to be be able to do is set the afsd memcache to
zero and find a way to work with the underlying OS'es buffer cache.