Date: Fri, 19 Sep 2008 17:33:30 +0200 From: "fluffles.net" <bsd@fluffles.net> To: Dieter <freebsd@sopwith.solgatos.com> Cc: freebsd-hardware@freebsd.org Subject: Re: Test: HighPoint RocketRaid 3120 PCIex1 2xSATA controller under FreeBSD 7.1-PRERELEASE Message-ID: <48D3C64A.5010207@fluffles.net> In-Reply-To: <200809190548.FAA08998@sopwith.solgatos.com> References: <200809190548.FAA08998@sopwith.solgatos.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Dieter wrote: >> If you get about the same scores with dd, try using a higher read-ahead >> (vfs.read_max value, set it to 32 for example). Also sometimes it's >> required to use a higher blocksize to get full potential, try: >> >> newfs -U -b 32768 /dev/<raid device> >> >> Warning: using 64KiB blocksize you risk hanging the system under heavy >> load (like 2 bonnies running simultaniously). >> > > What's this about 64KiB blocksize hanging the system? > Hang awhile then recover, or hang forever need a reboot? > Is this a RAID thing or are normal disks at risk? > See this, posted 2 years ago to freebsd-fs: http://freebsd.monkey.org/freebsd-fs/200610/msg00024.html Scott Long says the bug in vfs_bio.c is known but difficult to fix. They also argued that using a 64KiB blocksize should not be necessary. I have been getting slightly higher scores with 64KiB blocksize though, also with perfect alignment. 32KiB should be safe to use, i have not managed to crash the system under heavy I/O load. I'm curious if my suggestions gave you any better results, though. Regards, Veronica
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?48D3C64A.5010207>