Blogbench RAID benchmarks

Francois Tigeot ftigeot at wolfpond.org
Tue Jul 12 12:06:14 PDT 2011


On Tue, Jul 12, 2011 at 09:09:58AM -0700, Matthew Dillon wrote:
>     blogbench --iterations=200 ...  and you really need to include the
>     actual blogbench output.  The final results are worthless.

I had not enough time to do long benchmarks on the machine anymore.
Howewer, I can now use another one with only 4GB memory. How many iterations
do you think would be enough to do a good test with this quantity of RAM ?

>     Linear write activity.  I wonder why it isn't
>     clustering the I/O's, though.

That was newfs_hammer

>     Try setting vfs.hammer.double_buffer=1.

Will do in a future test.

> :   0    3 15.94  288  4.48   1  0 85  0 14
> :   0    3 16.00  620  9.69   1  0 79  0 20
> :   0    3 16.00  352  5.50   1  0 84  0 15
> 
>     This doesn't look good either.  It shouldn't degrade that much,
>     though it's a bit hard to tell with iostat because it doesn't print
>     the disk busy %.

What could I use to get meaningful data ?

-- 
Francois Tigeot





More information about the Kernel mailing list