[Gluster-devel] speed tests with gls3 fuse and tla 441

Anand Avati avati at zresearch.com
Thu Aug 9 18:06:22 UTC 2007


Harris/Nathan,
 I am looking into the issue. the performance seems to be a problem for
small files (which get passed on in one write() call to the filesystem). The
actual delay seems to be coming from either VFS/fuse. When write-behind
replies back 'written' that fast, there seems to be some kind of 'idling'
before getting further calls from fuse. The write-behind definitely seems to
make a positive difference for large files though. I am investigating this
deeper currently.

avati

2007/8/8, Nathan Allen Stratton <nathan at robotics.net>:
>
> On Wed, 8 Aug 2007, Harris Landgarten wrote:
>
> > I just ran some speed tests with tla 441 and fuse-2.7.0-gls3
>
> I am running tla 439, but haveing the same issues with write-behind and
> read-ahead. We are only talking about 134M....
>
> TLA-439
> write-behind    on
> read-ahead      on
>
> [root at vs0 share]# time cp /usr/bin/* /share/test/
> real    2m24.436s
> [root at vs0 share]# time cp -r /share/test/ /share/test2
> real    3m14.409s
> [root at vs0 share]# time rm -fr test*
> real    0m47.418s
>
> TLA-439
> write-behind    off
> read-ahead      on
>
> [root at vs0 /]# time cp /usr/bin/* /share/test/
> real    1m51.529s
> [root at vs0 /]# time cp -r /share/test/ /share/test2
> real    2m52.947s
> [root at vs0 /]# time rm -fr test*
> real    0m59.132s
>
> TLA-439
> write-behind    off
> read-ahead      off
>
> [root at vs0 share]# time cp /usr/bin/* /share/test/
> real    1m45.248s
> [root at vs0 share]# time cp -r /share/test/ /share/test2
> real    2m28.788s
> [root at vs0 share]# time rm -fr /share/test*
> real    1m1.128s
>
>
>
> ><>
> Nathan Stratton                         CTO, Voila IP Communications
> nathan at robotics.net                  nathan at voilaip.com
> http://www.robotics.net                 http://www.voilaip.com
>
>
>
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>



-- 
It always takes longer than you expect, even when you take into account
Hofstadter's Law.

-- Hofstadter's Law



More information about the Gluster-devel mailing list