A few days ago I posted a request asking what to do about disk blocks
that were unreadable. I got many (35+) replies and several requests
for a summary.
What to do about disk read errors can be broken up into two parts.
1) How to tell what file the error is is located.
2) How to keep the the error from happening again.
This line appeared on the console:
xy0g: read retry (hard ecc error) -- blk #149909, abs blk #258709
How to find which file a bad block resides is accomplished by "icheck
-b <block number within the partition> <raw device>" which will give the
inode number. So I did an "icheck -b 149909 /dev/rxy0g" and found the
inode to be a number like 1011. I did an "ncheck -i 1011 /dev/rxy0g",
and it gave me the file name. I had a couple of these to do, one
was in /usr/ucb/rsh and the other was not allocated to any file.
How to repair the bad spot is accomplished by running format's 'repair'
command to keep the block from being reallocated. Repair is supposed
to copy the contents of the block (if possible) and slip it so it will
never be used again. I did just that and just to be sure, I got
another copy of rsh and copied it on top of the existing one. With
repair it is VERY important to use the absolute block number.
After two days all is still okay.
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:05:57 CDT