Date: Sat, 14 Mar 2015 18:29:06 +1100 From: Peter Jeremy <peter@rulingia.com> To: Brett Wynkoop <freebsd-arm@wynn.com> Cc: freebsd-arm@freebsd.org, Mark Treacy <mark.treacy@gmail.com> Subject: Re: zfs Message-ID: <20150314072906.GA19582@server.rulingia.com> In-Reply-To: <20150313234510.3b41611b@ivory.wynn.com> References: <CAAdd%2BX5M5c%2BM0psLPBU40ihWvZkYQWa=NKFj%2B9HLAg3%2B7rpd4w@mail.gmail.com> <20150313234510.3b41611b@ivory.wynn.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--fUYQa+Pmc3FrFX/N Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On 2015-Mar-13 23:45:10 -0400, Brett Wynkoop <freebsd-arm@wynn.com> wrote: >While I think there would be a further space savings with dedupe I did >not want the system bogged down doing that work. The BeagleBone is >after all a pretty small system. Whilst an all-flash pool removes one of the main killers for dedupe, I suspect that the extra RAM to handle the DDT would outweigh the possible space-savings. >vfs.zfs.arc_max=3D"24M" =2E.. >With the above arc I am getting about 50% arc cache hits: Depending on your workload, you might find that increasing vfs.zfs.arc_meta_limit might help overall performance - at least on my x86 server, I find the default (25% of arc_max) a bit small. >I wonder if we could do zfs on root.....hmmmm. You'd need to build your own ubldr but it should be doable. >fast and without crashing the system. I also get the advantage of a >compressed filesystem and in addition to allowing me more storage it >may account for the better write performance. There are fewer bits to >write, so the slowness of flash media is somewhat compensated for. ZFS compression trades CPU resources for IO performance. For a modern x86 system, it's virtually always a win but the BBB has far less CPU available (though the flash is also very slow). >> Since zfs checksums everything you may well just be seeing it >> detecting corruption and repairing it. > >You could be right on this one. I think I need to dig out my other >Beagle Bone and shift the sd and usb flash over to it and run the >series of tests against ufs again on the new hardware. I say new >because I have never fired up the other BBone. Note that ZFS can't repair things unless you have redundant copies. Whilst there are redundant copies of metadata by default, there is only one copy of data - which is why you have: >errors: Permanent errors have been detected in the following files: > > /export/src/contrib/llvm/tools/clang/lib/Sema/SemaExpr.cpp > /export/ports/packages/All/perl5-5.18.4_11.txz > /export/ports/distfiles/rsync-3.1.1.tar.gz > /export/ports/distfiles/readline-6.3.tar.gz > /export/ports/comms/wsjt/files/configure You might like to set copies=3D2 if you don't have redundancy at the pool l= evel. > bbexport/ports:<0x253db> This is probably a lost metadata block. --=20 Peter Jeremy --fUYQa+Pmc3FrFX/N Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQJ8BAEBCgBmBQJVA+NCXxSAAAAAAC4AKGlzc3Vlci1mcHJAbm90YXRpb25zLm9w ZW5wZ3AuZmlmdGhob3JzZW1hbi5uZXRFRUIyOTg2QzMwNjcxRTc0RTY1QzIyN0Ux NkE1OTdBMEU0QTIwQjM0AAoJEBall6Dkogs0toQP/3HJQGNP74oC/ujDBctCPjPF 1ly5fwidCbA6XyxpzozGaist7Qb+94m8plCO7jGehohd2CqQprpwqD50ayn0Xjbs uBW8aLAvG1hSnQnNaqstW3v44ACeKrqc9sEoSYQYRBKlsNAsLnnVOhHTq09qXHFq yZ0NmUMcN2+DoD3IVP6N9d3JmNsHQfvCPihI6KsgUfUgIDWf394Fp7ehmq8lgpXo uAuDZuPXPcNns/UylazYRcSwMMbyrIe4vXKus7a2rEiCGA3CwZuikGJAlznjmNUP A8SMEO0HroGsTEK8LwMpr0S3OcA2YZBom5cuc8WxZ/haF3x6is3zUK+OXnqaOYZy rQz9wOFMkrxyUrT5qV3UxTuKhCB9X514ZC5bhcfGcY4dsEUwSPhu63JMrmk4l85U m7CtjfBZJC0GMibc+J0jMco/XT7YdtSFxVDcUDH5u0cUad3XG89G3RkZO9VfrdIa RBl53yjDqcjWQrcoZvec6EkXjUhgvHcniuRw/CCVgM2Nyn2aP+05ijERCOObQV7O O4Xik1XknaSYCoZV8yOEN9PlihAXMoDy4s1aPpLDFrpdfWj7VBD7JntrDm6hWFMI MOe1ermy72dCRtSZehlTJ+trklRMcxUOF8MoL6cVGnJSODRju9P+owvrssaoWSrH Q6FS89YFG32jQfErY6b7 =iy24 -----END PGP SIGNATURE----- --fUYQa+Pmc3FrFX/N--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20150314072906.GA19582>