Date: Wed, 19 Aug 2015 17:25:00 +0100 From: Gary Palmer <gpalmer@freebsd.org> To: PK1048 <paul@pk1048.com> Cc: FreeBSD Filesystems <freebsd-fs@freebsd.org>, javocado <javocado@gmail.com> Subject: Re: Optimizing performance with SLOG/L2ARC Message-ID: <20150819162500.GC13503@in-addr.com> In-Reply-To: <3FE10173-656C-4744-AB2D-32148A34CB46@pk1048.com> References: <CAP1HOmTidC3%2BG4XfhvkQxieo%2BSYMq-JWiXF9Cs4FSW2VqkktWA@mail.gmail.com> <023F881D-CCC5-4FCA-B09D-EB92C3BFBC03@pk1048.com> <CALeZrrRex43B6N7GkTtaP36v5q7mQF8U9V1bU2W1%2B2uh8fhhiQ@mail.gmail.com> <3FE10173-656C-4744-AB2D-32148A34CB46@pk1048.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Aug 19, 2015 at 12:10:39PM -0400, PK1048 wrote: > On Aug 19, 2015, at 12:02, Schweiss, Chip <chip@innovates.com> wrote: > > ZFS doesn't play well with the > > ZIL on the pool with SSDs. Even an SSD of the same type as the pool > > devices as the log device will fix the latency problem and throughput > > problems. > > If your load is sync writes then you decidedly want a LOG device, even if it is the same type as the devices in the pool. For the reasons others have posted. > One thing I am curious about: A lot of posters in the past (not blaming anyone in this thread) have said that the best way to find out if a SLOG device will help your application is to try. While ZFS gathers quite extensive statistics about read/write performance & volumes, ARC/L2ARC stats, etc, it doesn't seem to have any data about the number of writes to the pool that are sync vs async. It has to know, else it couldn't handle them properly. So why are there no stats about percentage of writes that are sync? Thanks, Gary
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20150819162500.GC13503>