On Fri, 26 Apr 1996, T H Pineapple wrote:

> [1] is it possible to use disks with netbsd-1.0 filesystems safely on
> a machine with a 1.1 kernel for read/write activity?
> 
	yup. tho if memory serves 1.1 has a clean flag for filesystems,
	so if you're swapping back to 1.0 and then to 1.1 again you might
	want to 'force' fsck to check if you're feeling paranoid.
	(swapping back and forrth shouldnt introduce any corruption, but
	if you boot 1.1 & shutdown cleanly, then boot 1.0 and crash the
	system the clean flag might be left set so next time you boot 1.1
	it wont check the filesystem, if you see what I mean :)
	Just 'fsck -f' form single user mode, or add -f to the fsck in
	/etc/rc until you've decided you dont need to boot 1.0 anymore.

	dont mix 1.0 system binaries (ie mountd, nfsd etc) with a 1.1 
	kernel or vica versa. Anything you've compiled yourself should be
	ok. (the interface to various kernel stuff changed 1.0 -> 1.1)

> [we may have to swap machines out due to the second problem, and
> it'll be a good chance to take the 1.0 box to 1.1 at last whilst
> still keeping the project drives serveavble]
> 
	Project drives should be fine 1.0 <-> 1.1

> [2] our main server drive [4 gig] for the lan is reporting more and
> more bad blocks - it used to be just one block constantly being
> reported on a single filesystem until i mapped it out with hdtoolbox
> from the Amiga side, backed up the partition, rebuild the filesystem
> and restored], but over the last 2 nights another 10 have come along.
> and they're on the photogenics 2 filesystem, which is
> mission-critical to say the least.
> 
> the drive's on an a3k with netbsd 1.0, btw, all disks are SCSI
> running on the in-built controller.  it looks like either cabling
> problems [the big disk is external with a DAT on 50-way, the cd-build
> drive and root/user/swap are internal on ribbon] or the drive itself
> is on the way out [it's cased, within warranty and has been reliable
> for a couple of years]- neither is good news, but workable after the
> deadline.
> 
> last time i backed up, mapped and rebuilt the entire fs - is is safe
> simply to delete the affected file, map the contained block out, fsck
> to sanity-check and then just keep using the filesystem?  [this could
> save a lot of time in the long run]
> 
	the former should be fine, but if you have a drive with creeping
	bad blocks get it swapped out as soon as you can...

> any pointers *really* appreciated on this one.  deadline time
> approaches, and we can't afford to lose any data and have drives
> explode at the 11th hour...
> 
> cheers.
> 
> thp / c!truS'96:   htthp://stupid.cdrom.work!.hit.hit.hit
> almathera UK's web/bbs/LANop. ie the only one mug enough to do it...
> 


		  David/abs 		(david@mono.org)

+44 181 888 8949 System Manager, Southern Studios Ltd, Box 59, London N22 1AR
 >=- OpenBSD http://www.openbsd.org. Free un*x for {sparc,i386,m68k,...} -=<
       <<< Monochrome - Largest UK Internet BBS - telnet://mono.org >>>