Date: Thu, 31 Dec 2009 16:36:53 -0600 (CST) From: Bob Friesenhahn <bfriesen@simple.dallas.tx.us> To: jhell <jhell@DataIX.net> Cc: freebsd-fs@freebsd.org Subject: Re: ZFS prefetch problem Message-ID: <alpine.GSO.2.01.0912311628190.1586@freddy.simplesystems.org> In-Reply-To: <alpine.BSF.2.00.0912311708440.72873@qvzrafvba.5c.ybpny> References: <ea7b7b810912310813x61af52acs5ecabd5615ce83d8@mail.gmail.com> <alpine.BSF.2.00.0912311708440.72873@qvzrafvba.5c.ybpny>
next in thread | previous in thread | raw e-mail | index | archive | help
This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. ---559023410-76287759-1262299013=:1586 Content-Type: TEXT/PLAIN; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 8BIT On Thu, 31 Dec 2009, jhell wrote: >> >> looks good. >> tested on 8-STABLE amd64, with/without log/cache ssd. (but not with >> zil_disable) >> > > Thanks for the heads up on this. I have applied the following to a stable/7 > system @r201336 "The obrien new year copyright ;)" and will write back with > further results. Now if someone can figure out how to crank up (double) the prefetch rate for my little test, I would be happy. 145MB/second is not nearly enough, especially if zfs can sometimes deliver 5X (or more) that rate. The read performance I see here (under Solaris 10) is very similar to the numbers posted by Mickaël Maillot. In fact initial read performance values posted by many others was almost the same, regardless of how many drives are available (e.g. 2 or 48 SAS drives). This suggests that zfs is not accellerating prefetch fast enough to make much difference for 8GB files. Bob -- Bob Friesenhahn bfriesen@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ ---559023410-76287759-1262299013=:1586--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.GSO.2.01.0912311628190.1586>