Debbie, the engineer who had worked on this problem set me straight. I
had posted second hand information to this list last week, and it wasn't
entirely correct (not even a little bit!) My apologies to Debbie,
Scott, and the Tru64 List. So, to ensure I don't mutilate the problem
and explanation again, I will use the cut n' paste method:
The problem with volrootmir had nothing to do with root. The problem
stemmed from the swap=dsk# option in the command. The 'disk write
failure' error will also only occur when the LSM simple disk is created
on the original swap disk on partition a. The simple disk contains the
metadata for rootdg and is created on the next available 'unused
partition'. Since swap was put on parition 'h', the next 'unused'
partition in your case was 'a'. Therefore when it tried to mirror the
swap disk, it attempted to write to block 0 when mirroring the simple
disk above.
The fix is to now check to see if this is the case and if so, create the
privoffset on the new simple disk to be what the original disk is
using.
The reason for the couple of days it took to receive this patch was due
to testing that needed to be accomplished first. Scott nor I wanted to
provide a patch that we were not positive would fix the
issue.
I hope this clears everything up.
Received on Tue Oct 05 1999 - 15:50:00 NZDT