[Top] [All Lists]

Re: "badblocks" for XFS?

To: "Jonathan F. Dill" <dill@xxxxxxxxxxxx>
Subject: Re: "badblocks" for XFS?
From: Eric Sandeen <sandeen@xxxxxxx>
Date: 13 May 2002 12:43:34 -0500
Cc: Stefan Smietanowski <stesmi@xxxxxxxxxx>, Mike Burger <mburger@xxxxxxxxxxxxxxxxx>, Linux XFS Mailing List <linux-xfs@xxxxxxxxxxx>
In-reply-to: <1021296035.1577.9.camel@xxxxxxxxxxxxxxxxxxxxx>
References: <Pine.LNX.4.44.0205130653090.3542-100000@xxxxxxxxxxxxxxxxxxxxxxxxx> <3CDFB2A0.9020700@xxxxxxxxxx> <1021296035.1577.9.camel@xxxxxxxxxxxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
On Mon, 2002-05-13 at 08:20, Jonathan F. Dill wrote:
> Is there any way to allocate/mark bad the badblocks with XFS so they
> won't be used by the filesystem?  SCSI drives usually have their own way
> to do this internally in the defects list, but AFAIK EIDE drives do not.

As I understand it, all modern drives do defect management internally,
remapping data blocks as they go bad.  If you're actually seeing a bad
block from the outside, that probably means that the drive has run out
of blocks to remap to, and it's all downhill from there.

At least, that's the reason I've always given for why filesystems
shouldn't track bad blocks - that's the drive's job.  If you want data
integrity, and your drive is showing bad blocks, throw it away and get a
new one.


> Historically, XFS grew up on SCSI drives with the IRIX "fx" utility to
> help add blocks to the drive's internal defect list, so I don't know if
> there is a really user-friendly, automated way to add bad blocks to an
> XFS filesystem on an EIDE drive.
> The only way I can see to do it is with xfs_db and xfs_bmap, although
> I'm not clear on the procedure.

Eric Sandeen      XFS for Linux     http://oss.sgi.com/projects/xfs
sandeen@xxxxxxx   SGI, Inc.

<Prev in Thread] Current Thread [Next in Thread>