From mboxrd@z Thu Jan 1 00:00:00 1970 From: NeilBrown Subject: Re: [md PATCH 00/16] hot-replace support for RAID4/5/6 Date: Fri, 23 Dec 2011 08:14:24 +1100 Message-ID: <20111223081424.682a1597@notabene.brown> References: <20111026014240.21110.28487.stgit@notabene.brown> <20111215171850.335da016@notabene.brown> <20111220161820.0830ef18@notabene.brown> <20111222215442.Horde.VHSHGZk8pphO85kS1Gs7yvA@cakebox.homeunix.net> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=PGP-SHA1; boundary="Sig_/svj7=jrtIi.j0OBx_iBvwXb"; protocol="application/pgp-signature" Return-path: In-Reply-To: <20111222215442.Horde.VHSHGZk8pphO85kS1Gs7yvA@cakebox.homeunix.net> Sender: linux-raid-owner@vger.kernel.org To: Alexander =?UTF-8?B?S8O8aG4=?= Cc: "Williams, Dan J" , linux-raid@vger.kernel.org List-Id: linux-raid.ids --Sig_/svj7=jrtIi.j0OBx_iBvwXb Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable On Thu, 22 Dec 2011 21:54:42 +0100 Alexander K=C3=BChn wrote: >=20 > Zitat von NeilBrown : >=20 > > On Wed, 14 Dec 2011 23:14:04 -0800 "Williams, Dan J" > > wrote: > > > >> On Wed, Dec 14, 2011 at 10:18 PM, NeilBrown wrote: > >> >> f248f8c md: create externally visible flags for supporting hot-repl= ace. > >> >> > >> >> 'replaceable' just strikes me as a confusing name as all devices are > >> >> nominally "replaceable", but whether you want it to be actively > >> >> replaced is a different consideration. =C2=A0What about "incumbent"= to mark > >> >> the disk as currently holding a position we want it to vacate and > >> >> remove any potential confusion with 'replacement'. >=20 > 'vacating' strikes me as the obvious choice, no? Interesting suggestion, but not quite the right meaning. The bit can be set before a replacement is available. So it isn't a statement about what is happening to the device, but about what should happening to the device. The meaning we want is "replace this device as soon as possible". I don't think there is one word that really has a meaning like that. "replaceable" was the closest I could get but that means "can be replaced" rather than "should be replaced" and as Dan pointed out that is a significa= nt difference. Maybe "deprecated"?? However I don't think using that would improve clarity. Thanks, NeilBrown --Sig_/svj7=jrtIi.j0OBx_iBvwXb Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.18 (GNU/Linux) iQIVAwUBTvOdsDnsnt1WYoG5AQJOMA//ZKO7phRsBpl2k38UZP6m35xe5VG+EGiw bpfO+BGvsrQARp4W3AZPZ/g2RYs10GPlLG51eqCuonPJy9+ZJ5ROjOqb6l7eG87S yYpmRNW85Dcqxcgvi0kn8kx7bkXQOxWblQL0+SuAiSnyUIz90+qs0q4CO4SqIv6v F25RsyqmePgTYYeCxr2R4uLhTsAXcHVrOjx8/XnGR4RtksIfRyB9XZ2nNOn5IzlA KzDzFDNypVRU1ryCaFMJqFzSkjFkTAt6dJ3/peJwdgNU0Dc/k767e3VJY58XTfFB 0fTm87TRm30rlfozL+19QeAhqJQMe+uqMJ9WbBBbhBwSS3h+5B6Zahq7eS3sNnsN jjO9vNIyZEikrPUfZUiiCEghtY5I1gX3R/ogcoUUk1ag7bP2xP+C1dketSnAmajP 3zOY0ePkFIS8U7FZ+EA94cCt1RbMCs3w0Avvs4dURhcRuzHgbLwpNlo1/uFiR5Nd mE/aX7ncBh1f0ngSvkxpFhY1VM+VSS2d7j/Y+txd0g212Dl8Q534c/17yRm3qF4U jnKFojDBBSwM7AjjIBA4W4pwO0D2RhLSVtAFetS4NVWCdq3TXwrA3WahhQBQ+iVU YGYD2h+JlGYgQn69Xhg+JeOH2/ogeK05WPSiRM9MqEn3qoZwidd7FnOi+TQ9INrs 56zacX2sMUs= =w/7c -----END PGP SIGNATURE----- --Sig_/svj7=jrtIi.j0OBx_iBvwXb--