Date: Wed, 6 Jan 2010 18:38:07 +0100 From: Ivan Voras <ivoras@freebsd.org> To: Alexander Motin <mav@freebsd.org> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r201658 - head/sbin/geom/class/stripe Message-ID: <9bbcef731001060938k2b0014a2m15eef911b9922b2c@mail.gmail.com> In-Reply-To: <201001061712.o06HCICF087127@svn.freebsd.org> References: <201001061712.o06HCICF087127@svn.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
2010/1/6 Alexander Motin <mav@freebsd.org>: > Author: mav > Date: Wed Jan =C2=A06 17:12:18 2010 > New Revision: 201658 > URL: http://svn.freebsd.org/changeset/base/201658 > > Log: > =C2=A0Increase default block size from 4K to 64K. It was reduces 6 yeard = ago, > =C2=A0when trees were big and FAST mode was enabled by default. > > =C2=A0So small block size doesn't benefits linear I/O operations in FAST = and > =C2=A0significantly slowdowns in ECONOMIC (default) mode. For single stre= am random > =C2=A0I/Os so small block doesn't give much benefits, as access time is u= sually > =C2=A0bigger then transfer time there. Same time it requires all heads to= seek > =C2=A0together for every single request, reducing performance on parallel= load. I think there was one more reason - though I'm not sure if it is still valid because of your current and future work - the MAXPHYS limitation. If MAXPHYS is 128k, with 64k stripes data was only to be read from maximum of 2 drives. With 4k stripes it would have been read from 128/4=3D32 drives, though I agree 4k is too low in any case nowadays. I usually choose 16k or 32k for my setups.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9bbcef731001060938k2b0014a2m15eef911b9922b2c>