[ltt-dev] performance results I measured with the latest lttng patches

KOSAKI Motohiro kosaki.motohiro at jp.fujitsu.com
Fri Sep 26 02:41:45 EDT 2008


> To to give an idea of the order of magniture of kernel effect on tbench
> on my dual 4-cores box Intel box :
> stock kernel 2.6.27-rc6 : 2024MB/s
> stock kernel 2.6.27-rc7 : 1989MB/s
> 1.7% slowdown, with exactly the same config. And this is between
> rc releases where the diff consists mostly of _bugfixes_.
> I start to doubt tbench represents well "real" workloads, given it's
> mostly doing kernel interaction and almost nothing in userspace.

In general, almost kernel developers think tbench is good real workload like benchmark.
it simulate smb file server (smb protocol enforce small many i/o).

In addition, some major regression tracking team (e.g. in Intel, IBM) use it
for i/o regression check.

If you think it isn't real workload, we should make altenative better benchmark.
Otherwise, LTTng is never accepted by performance monger.

At least, I'll try to improve LTTng before make alternative benchmark.
if other person also think the same, I'm glad.

More information about the lttng-dev mailing list