Disk Error

[Doug Hardie]

I have been getting the following disk errors consistently for the last 
month.

> ad2s1e: hard error reading fsbn 6934399 of 3467168-3467295 (ad2s1 bn 
> 6934399; cn 431 tn 164 sn 52) status=59 error=40
> spec_getpages:(#ad/0x20014) I/O read failure: (error=5) bp 0xc5678f94 
> vp 0xcb5f3a80
>                size: 65536, resid: 65536, a_count: 65536, valid: 0x0
>                nread: 0, reqpage: 0, pindex: 504, pcount: 16
> vm_fault: pager read error, pid 35441 (expireover)
>

How do you figure out which file has the problem?  expireover's logs 
are all buffered so you don't get the last partial buffer.  I don't 
know yet if I can mark that particular sector as bad, but if I can find 
the file I can at least move to someplace where it won't get deleted.  
I chased through the core dump and the only directory indicated but all 
of those files are good.  I have also tar'd the entire news directory 
elsewhere and no errors were encountered.  The sector is the same every 
day.