[CentOS] 2TB limit, weird mounting issues on reboot

[W S]

I'll read up more on LVM and GPT partitions.  Please forgive me for not
completely RTFM'ing this thing.  I was just at the end of my rope and
frustrated and figured someone on the mailing list would have some sage
advice.  And it appears that is the case. :)

Thank you to all who replied.

Regards,
Will

On 10/5/06, John Newbigin <jnewbigin at ict.swin.edu.au> wrote:
>
> You can not have a DOS partition larger than 2Tb.  parted will create
> one and tell the kernel about the size which works.... until you reboot
> and then the partition is 2Tb smaller than it should be because the
> partition table can not store the MSB of your partition size.
>
> The solution is to use LVM or GPT partitions.
>
> John.
>
> W S wrote:
>
> > Greetings,
> >
> > I'm seeing some weird issues surrounding the 2 terabyte limit on my
> > 3ware (9590SE, PCI Express model) raid controller.  Currently, the raid
> > array is 6 x 500GB disks for a total of around 2.3 terabytes formatted.
> > Looking at the documentation on the 3ware site, the raid card can handle
> > a single partition over 2TB, but it still looks like fdisk fails to
> > create a primary partition over the 2TB limit.  I did resort to parted,
> > and it creates a partition over 2TB.  And, in fact, I can run mke2fs on
> > this partition, and mount it.  I also have an entry for this partition
> > in the /etc/fstab file.
> >
> > However, if I reboot the system, the partition fails to mount from the
> > init scripts.  If I comment out the /etc/fstab line, and run a mount
> > command from rc.local, that works; however, the partition is now seen as
> > a 259GB partition, and not the 2.3TB I originally had to start.  I have
> > the feeling the partition table is getting hosed on a reboot.  Fdisk
> > seems to handle the partition on the raid array just fine if it is a
> > hair under 2TB, and from all the things I've read, it seems fdisk and
> > other disk utilities still have real problems with partitions greater
> > than 2TB (32-bit x 512-byte blocks, etc).
> >
> > I guess I'm shocked.  I figured this wouldn't be an issue as of 2006,
> > but it still is..... anyone have any similar experiences?  If so, any
> > workarounds other than just stay at the 2TB limit?
> >
> > Regards,
> > Will
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > CentOS mailing list
> > CentOS at centos.org
> > http://lists.centos.org/mailman/listinfo/centos
>
>
> --
> John Newbigin
> Computer Systems Officer
> Faculty of Information and Communication Technologies
> Swinburne University of Technology
> Melbourne, Australia
> http://www.ict.swin.edu.au/staff/jnewbigin
>
> _______________________________________________
> CentOS mailing list
> CentOS at centos.org
> http://lists.centos.org/mailman/listinfo/centos
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.centos.org/pipermail/centos/attachments/20061005/80bcf55a/attachment.html>