Date: Tue, 16 Oct 2007 04:43:23 -0700 From: Jeremy Chadwick <koitsu@FreeBSD.org> To: Peter Schuller <peter.schuller@infidyne.com> Cc: freebsd-current@freebsd.org Subject: Re: Status of MySQL on 7 w/o patches applied Message-ID: <20071016114323.GA36300@eos.sc1.parodius.com> In-Reply-To: <ff27tu$k01$1@ger.gmane.org> References: <20071016110642.GB54457@hyperion.scode.org> <ff27tu$k01$1@ger.gmane.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Oct 16, 2007 at 01:40:31PM +0200, Ivan Voras wrote: > Peter Schuller wrote: > > I ask because I recently (a couple of weeks ago) did comparisons using > > sysbench (comparable to what was used in the benchmark used previously > > in the report) between 6.2 and 7-CURRENT. Selects were actually half > > as fast or less than 6.2 in this particular case (identical hardware). > > Just in case you missed some debugging knobs: can you retest with > RELENG_7? (not -current). Also for Peter: can you try different schedulers? The stock RELENG_7 scheduler is 4BSD (reliable but not that fast), while ULE provides an *immense* improvement. The benchmarks I've seen (comparing old vs. new) in regards to MySQL were specifically in regards to the new ULE scheduler. The ULE scheduler in RELENG_7 is not the same as it was in RELENG_6. You'll need to rebuild your kernel for this. You can use the GENERIC kernel template and simply change SCHED_4BSD to SCHED_ULE. -- | Jeremy Chadwick jdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB |
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20071016114323.GA36300>