run fsck interactively

[Kirk MacLean_2]

Looking for corrective action for a suspected disk problem. I have listed below the symptoms and everything I could find about it. Thank you for taking a look.

Kmac-

This was the error listed prior to the "run fsck interactively screen". message:

/dev/dsk/c0t6d0 : Unallocated Inode has bad ic_contin value I=1917
owner=root mode=0
/dev/dsk/c0t6d0: size=0 Mtime=Dec 31 19:00 1969 /dev/dsk/c0t6d0: internal
error: GOT TO replay ().

HPUX 10.20 Workstation with an HFS file system.

Your IT resource center search for ic_contin produced no results!

In the struct inode declaration in the following appears:

ino_t ic_contin /* 124: continuation inode number */

[Michael Elleby III_1]

Hmmm, haven't seen the error before, but first, I'd do a pvdisplay on the disk to get a good idea if the disk is intact.

Then to run fsck interactively:

fsck -F hfs -o full 'full name of lv that sits on the disk' instead of the disk device.

HTH,

Mike-

[Michael Elleby III_1]

Kirk-

Just rethought my thinking, I don't think that fsck allows the use of the '-o full' option to use in the command as for hfs filesystems. I believe it would be:

fsck -F hfs /dev/rdsk/c0t6d0

Hopefully, this should work.. My apologies.

Mike-

[Michael Tully]

Don't fool around with it for too long you probably have a defective disk. Get a couple of good backups if you don't have disk/LV mirroring and get the disk replaced especially if it is under manitenance. Also make sure you have your ignite tapes up to date. Always keep ignite tapes up to date.

[Michael Steele_2]

Nice thing about HFS is the /etc/sbtab file. Find your file system in /etc/sbtab (* super block table *) and choose a number, say 1024. Then :

fsck -F hfs -b 1024 /dev/vg##/rlvol#

Super blocks are copies of the file system index and contains all the address pointers to locations of stored data blocks on the disk, and by using one of the copies you can recover the data.