Date: Sun, 10 Dec 2006 16:49:03 -0800 (PST) From: Matthew Jacob <mj@feral.com> To: Pawel Jakub Dawidek <pjd@FreeBSD.org> Cc: Perforce Change Reviews <perforce@FreeBSD.org>, Matt Jacob <mjacob@FreeBSD.org> Subject: Re: PERFORCE change 111375 for review Message-ID: <20061210164745.R58958@ns1.feral.com> In-Reply-To: <20061210231839.GE12193@garage.freebsd.pl> References: <200612100554.kBA5sJpn089774@repoman.freebsd.org> <20061210231839.GE12193@garage.freebsd.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
This was discussed at length some months back and the consensus with PHK and others was to do this at the GEOM level. If you don't mind, I'll look at yours. > On Sun, Dec 10, 2006 at 05:54:19AM +0000, Matt Jacob wrote: >> http://perforce.freebsd.org/chv.cgi?CH=111375 >> >> Change 111375 by mjacob@mjexp on 2006/12/10 05:53:25 >> >> Just to get it down... very first baby steps cut at the >> start of a multipath geometry class. > > Hmm, you are doing multipath GEOM class? > I already has a GEOM class which does dynamic multipathing and failover > at GEOM level. I thought this is incorrect way of implementing > multipathing. I think Scott was talking about doing it at the CAM > level AFAIR. > > Anyway, my code is in: > > //depot/user/pjd/geom_classes/sys/geom/rome/... > > -- > Pawel Jakub Dawidek http://www.wheel.pl > pjd@FreeBSD.org http://www.FreeBSD.org > FreeBSD committer Am I Evil? Yes, I Am! >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20061210164745.R58958>