sjb@dalek.UUCP (Seth J. Bradley) (09/15/88)
I'd like to share some experiences I had last weekend in case others have problems similar to mine. I recently observed massive corruption on my second drive (which I use as /usr, /dev/dsk/1s2). It is a Seagate ST251. After I got everything back, I did an fsck and got disk errors and unreadable blocks. I did a backup and then bought Speedstor to check out the disk. This uncovered the sector that was causing the hard errors, but after locking it out I still couldn't get fsck to pass the new partition. I kept gettting cannot read or cannot write errors. I finally found the problem. I had been using divvy to determine the size of the partition. If you have a single partition, divvy will assign every available block on the disk to the partition. I remembered something about System V needing the 1200 highest numbered blocks for "housekeeping". So I did a new mkfs with the number of blocks reduced by 1200, and everything went fine! I think Microport should change the divvy program. P.S. I lost two binaries, /usr/bin/checkeq and /usr/lib/lpget when I did my backup. Could anyone out there (especially Mport people) tell me which floppies to extract these from and the necessary cpio command? Thanks in advance! -- Seth J. Bradley UUCP: uunet!lll-winken!dalek!sjb Internet: lll-winken.llnl.gov!dalek!sjb