linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Possible addition to contest
@ 2002-09-16 18:29 Bill Davidsen
  2002-09-16 20:37 ` Con Kolivas
  0 siblings, 1 reply; 2+ messages in thread
From: Bill Davidsen @ 2002-09-16 18:29 UTC (permalink / raw)
  To: contest; +Cc: Linux-Kernel Mailing List

It would be nice to be able to add a short string to the kernel version
for identification, to differentiate between runs with the same kernel and
different tuning.

I've found that tuning often helps response, like reducing memory with
older rmap versions (ran for a while using mem=256m), or tuning bdflush
with -aa kernels, which I did after Andrea gave me some serious hints.
Just a short string added to the version would make this a bit easier to
follow after the fact.

-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: Possible addition to contest
  2002-09-16 18:29 Possible addition to contest Bill Davidsen
@ 2002-09-16 20:37 ` Con Kolivas
  0 siblings, 0 replies; 2+ messages in thread
From: Con Kolivas @ 2002-09-16 20:37 UTC (permalink / raw)
  To: Bill Davidsen; +Cc: Linux-Kernel Mailing List

Quoting Bill Davidsen <davidsen@tmr.com>:

> It would be nice to be able to add a short string to the kernel version
> for identification, to differentiate between runs with the same kernel and
> different tuning.

Easy fix. Will do. The original version you had to tell it which kernel you were
running so I can add that as optional.

> I've found that tuning often helps response, like reducing memory with
> older rmap versions (ran for a while using mem=256m), or tuning bdflush
> with -aa kernels, which I did after Andrea gave me some serious hints.
> Just a short string added to the version would make this a bit easier to
> follow after the fact.

Perfectly reasonable. Be aware though that reducing mem will also change the
write size of io loads in contest. Although this will test the VM's response to
half and full size writes it will also decrease the size of IO load.

Con.


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2002-09-16 20:32 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-09-16 18:29 Possible addition to contest Bill Davidsen
2002-09-16 20:37 ` Con Kolivas

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).