RE: drive failure, now 'cannot alloc 494581644 bytes for inoinfo' and 'bad inode number 3556352 to nextinode'

> -----Original Message-----
> From: Doug White [mailto:dwhite_at_gumbysoft.com]
> Sent: Tuesday, April 06, 2004 9:52 PM
> To: Will Saxon
> Cc: current_at_freebsd.org
> Subject: Re: drive failure, now 'cannot alloc 494581644 bytes for
> inoinfo' and 'bad inode number 3556352 to nextinode'
> 
> 
> I can pretty much assure you the volume isn't OK. The types of errors
> you're seeing are indicative of severe corruption, usually 
> due to random
> data being written over critical filesystem blocks.  I'd 
> suggest running a
> parity verify against the volume to force corrections to start with --
> this can't make it any worse than it already is, and may recover the
> damaged blocks on the disk that lost power.
> 

Is there a freebsd tool that will let me do the parity verify? This
controller has the most minimal BIOS interface possible, I think they
want all the real work to be done through a windows utility.

> You could also try running fsck against an alternate superblock to get
> around any corruption thats specific to the primary superblock, but my
> experience with this kind of failure has shown that there's 
> usually more
> significant damage than just a mulched superblock.
> 
> Its quite likely that the filesystem is not recoverable.   
> Hope you have
> backups :-)

Not quite, but it was 'in testing' anyway. I guess it failed the test!

-Will