Date: Sun, 9 Jun 2002 10:56:57 -0700 From: "Philip J. Koenig" <pjklist@ekahuna.com> To: Stephen Hovey <shovey@buffnet.net> Cc: questions@FreeBSD.ORG Subject: Re: Filesystem paradox, take 2 Message-ID: <20020609175656821.AAA574@empty1.ekahuna.com@pc02.ekahuna.com> In-Reply-To: <Pine.BSF.4.05.10206090923260.10468-100000@buffnet11.buffnet.net> References: <20020609050258756.AAA535@empty1.ekahuna.com@pc02.ekahuna.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 9 Jun 2002, at 9:26, Stephen Hovey boldly uttered: > > to mount: > > > > mountroot> ufs:/dev/amrd0s2a > > > > ..it mounts root read-only because it tells me the filesystem is > > dirty so I have to run fsck. But!! fsck won't run, because it > > depends on /etc/fstab to tell it what filesystems to check, and I see > > no way to manually specify a device! AND, I can't edit fstab, > > because it's on / and / is mounted read-only! Argh! > > You can specify what to fsck - I believe its just > fsck /dev/ramrd0s1a (the raw one not the block one) Apparently you're right, I just tested this on another box. I was certain I tried this on the original machine and it didn't work... hmm. (also the manpage for fsck uses the term "filesystem" in the command syntax.. I always parsed that to mean something like "/var", rather than the raw device ie "/dev/da0s1e"..) > > PS: my next try is going to be deleting the FreeBSD slice from the > > RAID logical drive, boot off fixit CD, create a new FreeBSD slice, > > use disklabel to re-create all the filesystems, and copy/restore it > > in the conventional way - ie with a dump/restore or tar/untar pipe. > > But I sure wish I didn't have to go through all that.. > > The cleanest way, when all the hardware has changed (not only diff > controllers but you went to raid) is ALWAYS to restore from a backup of > the upper layer because the lower layer stuff (in my experience anyway) > will always mess ya up - if you approached it that way, then you would > have 'to go through all that..' that you are goin thru tryin to do a > direct image! You'd be done already and havin a beer! Yep, well it did work that way -- although I guess I've been spoiled by how easy it is to image-copy Windows boxes. The reason I spent so much time on it this time was I looked at it as a learning experience. BTW, the transfer rate when using the DOS imaging tools (DriveImage) was around 500MB/min, and using dump | restore with 64kb blocksize was about 1.2MB/min. This was mitigated somewhat by the fact that since DriveImage doesn't natively understand UFS partitions, it had to copy every sector. (It does understand Linux ext2 partitions, and I've suggested to them in the past to consider this for FreeBSD, but nothing yet) Thanks again, Phil -- Philip J. Koenig pjklist@ekahuna.com Electric Kahuna Systems -- Computers & Communications for the New Millenium To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020609175656821.AAA574>