Date: Fri, 16 Sep 2011 15:35:54 +0200 From: Terje Elde <terje@elde.net> To: "freebsd@top-consulting.net" <freebsd@top-consulting.net> Cc: "freebsd-questions@freebsd.org" <freebsd-questions@freebsd.org> Subject: Re: FS of choice for max random iops ( Maildir ) Message-ID: <32990703-D068-4B0D-AF3A-C1E6EA0A4100@elde.net> In-Reply-To: <20110916063153.200375qdq59crf8c@mail.top-consulting.net> References: <20110916063153.200375qdq59crf8c@mail.top-consulting.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On 16. sep. 2011, at 12:31, freebsd@top-consulting.net wrote: > Right now I defined an entire array of 8TB ( all 16 disks ) separated in t= wo pieces. 50 GB for FreeBSD to boot and the rest available to configure as s= torage. ZFS will want to write to it's ZIL (zfs intent log) before writing to the fi= nal location of the data. Even if you're not waiting for the ZIL-write to di= sk (because of the controller ram), those writes will probably make it throu= gh to disk. That gives you twice as many writes to disk, and a lot more seek= .=20 If you want to take zfs for a proper spin, I'd like to sugget adding two sma= ll SSDs to the setup, mirrored by zfs. You can use those both for the ZIL, a= nd also as cache, for the array. That's a fairly small investment these days= , and I would be surprised if it didn't significantly improve performance, b= oth for your benchmark, and real load.=20 Note: you might be in trouble if you loose your ZIL, thus the doubling up. I= *think* you can SSD a cache without risking dataloss, but don't take my wor= d for it.=20 Terje=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?32990703-D068-4B0D-AF3A-C1E6EA0A4100>