From mboxrd@z Thu Jan 1 00:00:00 1970 From: Loic Dachary Subject: controlling which OSD goes to which PG / testing a fix Date: Mon, 11 May 2015 13:51:08 +0200 Message-ID: <555097AC.6000000@dachary.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="WKSKPSoSQfwDr8cnsjba3XCx7upgXj5hf" Return-path: Received: from mail2.dachary.org ([91.121.57.175]:32861 "EHLO smtp.dmail.dachary.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753179AbbEKLvL (ORCPT ); Mon, 11 May 2015 07:51:11 -0400 Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Shylesh Kumar Cc: Ceph Development This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --WKSKPSoSQfwDr8cnsjba3XCx7upgXj5hf Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Hi, Sorry I did not get a chance to answer in time to your question on IRC.=20 loicd: a ruleset having {step take osd.0 step emit } should choose osd.0 for all pgs right=20 wats the error with this crushmap ruleset http://pastebin.com/y= Xn1t4sU i want all pgs to be on [0,1,3] which is working as expected but if 3 goesdown i want it to be [0,1,4] .. this is not happen= ing=20 shylesh: hi ! shylesh: looking at http://pastebin.com/yXn1t4sU (answering here = because that's the channel where non dev questions are best archived) shylesh: that's a very unusual crush map. You want to precisely c= ontrol the order in which the OSDs are assigned to a placement group, is = that right ?=20 If that's so, maybe your use case would be better served by other= means. What are you trying to achieve ? Measure the bandwidth between OS= Ds during repair maybe ?=20 loicd: I want one group of PG mapped to [0,1,3] and another gro= up of pgs(may be from different pool) mapped to [0,1,2] and I will mark 3= as out and 2 as down=20 so that new group becomes [0,1,4] , osd4 will have degraded and= misplaced pgs to get mapped there is a fix i want to test , degraded pgs should get priorit= y over misplaced pgs so now i want to see on osd4 degraded will be recovered first a= nd then misplaced now the mapping is working fine but if suppose 2 goes down i wi= ll be left with [0,1] for some reason 4 is not joining the replica group Could you show the fix that you want to test ? I may be able to figure ou= t a simpler strategy to verify that it does what is expected. Cheers --=20 Lo=C3=AFc Dachary, Artisan Logiciel Libre --WKSKPSoSQfwDr8cnsjba3XCx7upgXj5hf 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) iEYEARECAAYFAlVQl6wACgkQ8dLMyEl6F21xJACbBq9EG5O3uBdWkxABywmrGqnC giQAoIcu4p0lBJ3UN8NvUgYWUZuc6+4D =5DSe -----END PGP SIGNATURE----- --WKSKPSoSQfwDr8cnsjba3XCx7upgXj5hf--