Re: kern/55362: bad inodes created after dump|restore

[Patrick Welche <prlw1%cam.ac.uk@localhost>]

The following reply was made to PR kern/55362; it has been noted by GNATS.

From: Patrick Welche <prlw1%cam.ac.uk@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/55362: bad inodes created after dump|restore
Date: Wed, 10 Jun 2020 15:33:10 +0100

 On Wed, Jun 10, 2020 at 02:15:02PM +0000, Martin Husemann wrote:
 > The following reply was made to PR kern/55362; it has been noted by GNATS.
 > 
 > From: Martin Husemann <martin%duskware.de@localhost>
 > To: gnats-bugs%netbsd.org@localhost
 > Cc: 
 > Subject: Re: kern/55362: bad inodes created after dump|restore
 > Date: Wed, 10 Jun 2020 16:10:58 +0200
 > 
 >  I have seen such fallout with dying disks + raid, where writes to one
 >  sector did not read back the same (e.g. when both disks in a raid1 did
 >  actually write different things to the same sector and bad luck makes
 >  the raid return one or the other).
 >  
 >  Dump/restore is probably no factor in this at all.
 >  
 >  Can you try recreating the issue on a non-raid partition?
 >  Do we have some disk verification tool in pkgsrc?
 
 If it were the disks, could I expect
 - to have seen a similar problem on the zfs partition of those disks
 - for MLH not to have had a similar observation with a >=  9.99.64 kernel
 - for disks that are only a couple of months old and have been
   reliable to not suddenly mechanically behave strangely
 
 OK, tell me which hoops to jump through...