Date: Wed, 2 Dec 2009 08:00:19 -0500 From: Maxim Khitrov <mkhitrov@gmail.com> To: Randi Harper <randi@freebsd.org> Cc: Jerry McAllister <jerrymc@msu.edu>, Peggy Wilkins <enlil65@gmail.com>, freebsd-questions@freebsd.org Subject: Re: 8.0-RELEASE and "dangerously dedicated" disks Message-ID: <26ddd1750912020500y433dad1el83f7e8ed4b8e21ee@mail.gmail.com> In-Reply-To: <e277d6c80912012034u4ba1a6fdj97d72f4605bb3e0c@mail.gmail.com> References: <1789c2360911280928t1e6e7b06p707abc1131f82bef@mail.gmail.com> <26ddd1750912010459weda306esbb81d7a2b7025b6e@mail.gmail.com> <20091201210321.GA39958@gizmo.acns.msu.edu> <e277d6c80912012034u4ba1a6fdj97d72f4605bb3e0c@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Dec 1, 2009 at 11:34 PM, Randi Harper <randi@freebsd.org> wrote: > I'm going to just reply to all of these at once. > > On Tue, Dec 1, 2009 at 1:03 PM, Jerry McAllister <jerrymc@msu.edu> wrote: >> On Tue, Dec 01, 2009 at 07:59:42AM -0500, Maxim Khitrov wrote: >> >>> On Sat, Nov 28, 2009 at 12:28 PM, Peggy Wilkins <enlil65@gmail.com> wro= te: >>> > Due to history I won't go into, all my production (currently >>> > 7.2-RELEASE) systems are installed onto "dangerously dedicated" disks= . >>> > =C2=A0What exactly do I need to do to upgrade them to 8.0? =C2=A0(I'm= not asking >>> > for an upgrade procedure, I'm familiar with that, but rather, how thi= s >>> > change impacts the upgrade.) =C2=A0I think that the suggestion that t= he >>> > disks need to be reformatted is extreme and I hope something less >>> > extreme will suffice. > > > Just to point out the obvious, you shouldn't use "dangerous" and > "production" in the same sentence. :) > That depends on why "dangerous" was put into the name. In this case, it's a flag to indicate that we should understand what's going on underneath before using the feature in question. > >>> > Also, just to be clear, does this statement refer to boot disks, data >>> > disks, or both? >>> > >>> > It doesn't make sense to me that "dangerously dedicated" could have a= n >>> > impact on UFS filesystems specifically. =C2=A0A partition table is ju= st a >>> > partition table, regardless of what filesystems might be written on >>> > disks, yes? =C2=A0Am I misunderstanding something here? >> >> I don't know why it would have an affect, but they say it does. > > > Did you see all the mailing list chatter about new installations > failing due to sysinstall not being able to newfs device names that > didn't exist? This is related. Also, a partition table isn't just a > partition table. It's a little more complex than that. It has > *nothing* to do with the filesystems inside. It has everything to do > with the way that FreeBSD looks at the drive to figure out what's on > it. See man pages for geom/gpart. There are others that have given a > better explanation than I can provide (marcus, juli). Search the > archives. Trust me, I didn't remove DD support from sysinstall just to > make life more complicated for everyone. I did this because as it > stands right now, it doesn't work. > <snip> > > It's not a filesystem thing. See above. > > -- randi > I think this is where the misunderstanding is. Based on what you said, the documentation should read "dangerously dedicated mode is no longer supported in sysinstall," is that correct? If we don't use sysinstall for anything, the change doesn't affect us. The current wording leads people to believe that something was changed in the FreeBSD geom internals that would, for example, prevent DD disks from being recognized in 8.0. As I read geom(4), there is nothing new in the tasting section about the way devices are offered to geom classes. - Max
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?26ddd1750912020500y433dad1el83f7e8ed4b8e21ee>