Hello,

I got myself in a bit of trouble. My Multia running NetBSD-post-1.2.1
got `suddenly' one or two bad blocks on its internal SCSI disk. I still
have a fully functional machine as it boots from an external disk, but
I'd prefer not to put the internal one in the Hall of Fame of things
that used to serve well. So, does and how does NetBSD (-alpha) cope
with bad blocks. I found the badsect and bad144 commands, but these
seem either not what I need or refuse to actually do it. badsect
complains about the actual sectornumber being critical to the filesystem
and refuses to get rid of it. fsck reports the bad block but does not
list it as bad. Any pointers, ideas etc. I may be able to put the disk
in an Intel box (with NetBSD) if that's needed or helpful. However, I'd 
like to know how to handle this on the alpha, if possible. The partition
(the whole disk if needed) with the bad block is empty now (newfs-d), 
so I can implement drastic measures if required. Is there some low-level
scsi-formatting tool or so I could use???

Thanks,

Feico.

PS: These bad blocks probably are caused by the instable/shaky powerlines
to the Multia. Some of the powerlines are a bit torn by opening the box
too fast/rough once, and now my Multia suffers from spontaneous death 
syndrome. Shaking the powerlines a bit, revives it most of the time.
I'd like to repair this. So if any of you, knows whether it's easy and how
to replace the powerlines (from powersupply to motherboard) or where
I can get a cheap spare power-supply, I'd be happy to hear it.