Thanks to Stiffler Durden and alan_at_nabeth.cxo.dec.com for their
responses.
Alan suggested that the error log be checked for i/o errors. Stiffler
commented that bmt is the bitfile metadata table, which is part avfs on
disk metadata and further suggested that there was some domain
corruption due to hardware failure.
A check of the error logs did indicate that /dev/rz2 had logged a number
of CAM-SCSI events (record code 199). The drive was changed and
presently all is well.
Thanks again
frank
...............original post:
I am getting the following message (on an Alpha 8400) when I run the
verify on a particular domain:
check_disk: /dev/rz2g bad bmt read
check_disk: /dev/rz2h bad bmt read
What could be causing this and does it spell an impending crash ( the
system is still up)? Is there anything I can do to correct the
situation?
Received on Fri Sep 21 2001 - 11:47:06 NZST