Re: [Fwd: [BENCHMARK] 2.5.73, -bk1, -bk2, -mm1 regression results]

[Lou Langholtz <ldl@aros.net>]

Mark Peloquin wrote:

> Nightly Regression Summary
> for
> 2.5.72 vs 2.5.73
>
>
> Benchmark         Pass/Fail   Improvements   Regressions       
> Results       Results   Summary
> ---------------   ---------   ------------   -----------   
> -----------   -----------   -------
>
> dbench.ext2           P            N              N             
> 2.5.72        2.5.73    report
> dbench.ext3           P            Y              N             
> 2.5.72        2.5.73    report
> dbench.jfs            P            N              N             
> 2.5.72        2.5.73    report
> dbench.reiser         P            N              N             
> 2.5.72        2.5.73    report
> dbench.xfs            P            N              N             
> 2.5.72        2.5.73    report
> kernbench             P            N              N             
> 2.5.72        2.5.73    report
> lmbench               P            Y              Y             
> 2.5.72        2.5.73    report
> rawiobench            P            N              Y             
> 2.5.72        2.5.73    report
> specjbb               P            Y              Y             
> 2.5.72        2.5.73    report
> specsdet              P            N              Y             
> 2.5.72        2.5.73    report
> tbench                P            Y              N             
> 2.5.72        2.5.73    report
> tiobench.ext2         P            N              N             
> 2.5.72        2.5.73    report
> tiobench.ext3         P            Y              Y             
> 2.5.72        2.5.73    report
> tiobench.jfs          P            N              N             
> 2.5.72        2.5.73    report
> tiobench.reiser       P            Y              Y             
> 2.5.72        2.5.73    report
> tiobench.xfs          P            N              Y             
> 2.5.72        2.5.73    report
> volanomark            P            N              N             
> 2.5.72        2.5.73    report
>
> http://ltcperf.ncsa.uiuc.edu/data/2.5.73/2.5.72-vs-2.5.73/
>
> Nightly Regression Summary
> for . . .

Wow! This is impressive. I'd love to see this include performance of 
ext3 (or jfs) over NBD (network block device) too. Or at least 
performance metrics like this for NBD some other way. I've done some 
benchmarking myself for the work I've been doing on NBD but nothing near 
this comprehensive. If someone had the spare time to do this though for 
NBD, it'd help me enormously. Any interest??? :-)