xfs
[Top] [All Lists]

Re: XFS as Root filesystem

To: Russell Cattelan <cattelan@xxxxxxxxxxx>
Subject: Re: XFS as Root filesystem
From: "Stephen C. Tweedie" <sct@xxxxxxxxxx>
Date: Thu, 6 Apr 2000 17:23:50 +0100
Cc: "Stephen C. Tweedie" <sct@xxxxxxxxxx>, Lyle Seaman <lws@xxxxxxxxxxxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <38ECB5E1.E6CFEC35@xxxxxxxxxxx>; from cattelan@xxxxxxxxxxx on Thu, Apr 06, 2000 at 11:05:53AM -0500
References: <200004051440.JAA25697@xxxxxxxxxxxxxxxxxxxx> <38EB8FEF.9CB411C@xxxxxxxxxxxxxxxx> <20000406162144.B14727@xxxxxxxxxx> <38ECB5E1.E6CFEC35@xxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
Hi,

On Thu, Apr 06, 2000 at 11:05:53AM -0500, Russell Cattelan wrote:
> 
> XFS doesn't actually try to detect file system inconstancies at mount
> time.
> (this would defeat quick recovery)

Neither does ext3.  What the filesystem can do is to detect 
inconsistencies at run time (eg. clearing a bit in an allocation
bitmap but finding it already clear, or finding illegal block mappings
in a file --- that sort of thing), and record in the superblock that 
Bad Things have happened and a full consistency check is needed on 
reboot.  Only data corruption errors can lead to this.

> If at any time a live XFS file system detects and error XFS will
> immediately halt
> all IO to the device. It is then up to the administrator to do manual
> checks and
> or repairs.

Right, ext3 does too (in fact it gives you a per-fs option either to 
suspend IO on error, or to do an immediate panic).  In either case, 
though, it marks the presence of the error in the superblock so that
after a reboot, the system knows that forced fsck is required.

--Stephen

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