Messages in this thread | | | Date | Thu, 4 Sep 2008 13:35:29 +0200 | From | Ingo Molnar <> | Subject | Re: oltp ~10% regression with 2.6.27-rc5 on stoakley machine |
| |
* Lin Ming <ming.m.lin@intel.com> wrote:
> A huge trace result file, see the attachment for more.
> # tracer: sched_switch > # > # TASK-PID CPU# TIMESTAMP FUNCTION > # | | | | | > mysqld-3799 [07] 151.411009: 3799:120:R ==> 3807:120:R ... > sysbench-3803 [07] 151.417048: 3803:120:R + 3791:120:S
ok, this only spans 6 milliseconds, which is probably a bit too light.
Could you please increase the trace size 32-fold (via /debug/tracing/trace_entries), and send us the (large!) trace file compressed via bzip2 -9, off-list (or via a web link)? [anyone on-list can have a look at the smaller trace file already]
Thanks,
Ingo
| |