From mboxrd@z Thu Jan 1 00:00:00 1970 From: Loic Dachary Subject: Re: BlaumRoth with w=7 : what are the consequences ? Date: Thu, 25 Sep 2014 01:06:08 +0200 Message-ID: <54234E60.9070108@dachary.org> References: <5422F67B.3000302@dachary.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="tepm9nAV6hmmD3j6fpwE0Lku7jjXwGHXK" Return-path: Received: from mail2.dachary.org ([91.121.57.175]:59356 "EHLO smtp.dmail.dachary.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1750914AbaIXXGL (ORCPT ); Wed, 24 Sep 2014 19:06:11 -0400 In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Kevin Greenan Cc: Ceph Development This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --tepm9nAV6hmmD3j6fpwE0Lku7jjXwGHXK Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Hi Kevin, On 24/09/2014 20:40, Kevin Greenan wrote:> The constraint guarantees the = MDS property. I believe there are conditions where w+1 is composite and = you still have an MDS code, but there are restrictions on 'n' (codeword l= ength). So, you may have chosen the right parameters. Did you verify al= l possible combinations of erasures is tolerated? I tried all combinations that are likely to have been used and they all w= ork out. Here is the script I used: for w in 7 11 13 17 19 ; do for k in $(seq 2 $w) ; do for m in $(seq 1 $k= ) ; do for erasures in $(seq 1 $m) ; do ./ceph_erasure_code_benchmark --p= lugin jerasure --workload decoded --iterations 1 --size 4096 --erasures $= erasures --parameter w=3D$w --parameter k=3D$k --parameter m=3D2 --parame= ter technique=3Dblaum_roth ; done ; done ; done ; done does that mean we're safe despite the fact that w+1 is not prime in all s= ettings ?=20 > For sake of safety, you probably do not want to get experimental with '= production' code. IMHO, you should put the check in, especially if 'n' i= s tunable. n is tunable and the check is now enforced. I'm worried about the content= that was previously encoded with codeword length that do not match prime= (codeword + 1) Cheers --=20 Lo=C3=AFc Dachary, Artisan Logiciel Libre --tepm9nAV6hmmD3j6fpwE0Lku7jjXwGHXK Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iEYEARECAAYFAlQjTmAACgkQ8dLMyEl6F21vcgCgwv5zAEF/HLa5kAfF01yi2rxI aP8AoMHhjO9Erz1kyDWG8B4Yitw2D5WI =tJks -----END PGP SIGNATURE----- --tepm9nAV6hmmD3j6fpwE0Lku7jjXwGHXK--