[Top] [All Lists]


To: "Amit Agrawal, Noida" <amitag@xxxxxxxxxxxxxxxxx>
Subject: Re:
From: James Pearson <james-p@xxxxxxxxxxxxxxxxxx>
Date: Wed, 14 Aug 2002 10:58:04 +0100
Cc: Seth Mos <knuffie@xxxxxxxxx>, linux-xfs@xxxxxxxxxxx
Organization: Moving Picture Company
References: <E04CF3F88ACBD5119EFE00508BBB212103F20E54@xxxxxxxxxxxxxxxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
We've recently been trying out Medea (formerly Storage Concepts)
Fibreblocks which allow you to set the block size to 512 or 2048 bytes.

We made an XFS file system using the 2048 byte block size on the raid -
but we got lots of SCSI disk errors when reading/writing to the file
system - eventually the file system shuts down and we get a kernel oops
on umount.

ext2/3 file systems worked fine with the 2048 byte block size.

James Pearson

"Amit Agrawal, Noida" wrote:
> No I mean sector sizes which are multiple of 512 i.e 1024 or 4096,  of a
> logical device configured through volume management software.
> -----Original Message-----
> From: Seth Mos [mailto:knuffie@xxxxxxxxx]
> Sent: Wednesday, August 14, 2002 3:05 PM
> To: Amit Agrawal, Noida; linux-xfs@xxxxxxxxxxx
> Subject: Re:
> At 14:47 14-8-2002 +0530, Amit  Agrawal, Noida wrote:
> >Could you please tell me whether XFS will work with logical device having
> >sector size greater than 512 bytes or not?
> Tough question, I am not sure but something might break.
> So you mean you have a device that has sectors of say 520 bytes instead of
> 512 like some of the SAN appliances do?
> Cheers
> --
> Seth
> It might just be your lucky day, if you only knew.

<Prev in Thread] Current Thread [Next in Thread>
  • RE:, Amit Agrawal, Noida
    • Re:, James Pearson <=