Soft-Raid Testing and other stuff...

[Richard Hobbs]

Hello,

> -----Original Message-----
> From: redhat-list-bounces at redhat.com 
> [mailto:redhat-list-bounces at redhat.com] On Behalf Of Ed Wilts
> Sent: 20 January 2005 12:53
> To: General Red Hat Linux discussion list
> Subject: Re: Soft-Raid Testing and other stuff...
> 
> On Thu, Jan 20, 2005 at 11:17:01AM -0000, Richard Hobbs wrote:
> > > > I didn't duplicate the MBR as far as I know... I 
> assumed that RAID 1
> > > > (mirroring) mirrored the entire disk, sector for sector...
> 
> Nope - it's partition by partition.  That's why you did the raidhotadd
> *per partition*.
> 
> There are pros and cons to this approach - the biggest pro is that you
> don't have to mirror the entire disk.  The biggest con is 
> that only data
> partitions are mirrored.
> 
> For details on mirror the mbr, see:
> http://www.dur.ac.uk/a.d.stribblehill/mirrored_grub.html
> 
> I haven't tried this but it's referenced by the online Red Hat
> documentation.
> 
> > > > Also, if I then insert a blank disk, will it automatically 
> > > > re-generate the
> > > > data onto it? Will I have to duplicate the MBR manually again?
> > > > 
> 
> You will always have to do this manually.  Software RAID knows nothing
> about the MBR - only partitions.

So... am I correct in saying that if I run 'raidhotadd' on all partitions to
get them mirrored correctly, and then I run the grub stuff to install the
slave MBR, all is set and there's absolutely nothing else that needs to be
done? In order to have a fully resilient system?

I realise the BIOS might need to be altered to tell it to boot from the
slave if the master is dead, but apart from that, is everything done that
needs to be?

Also, what are the exact steps to follow if the master disk dies in a
soft-raid 1 environment?
What are the steps to follow if the secondary disk dies in a soft-raid 1
environment?

I assume they are fairly similar... Just replace the dead disk with an
identical blank one, boot up and reinstall the MBR once the automatic
reconstruction is done, right?? Or is it not that simple...

Thanks again,
Hobbs.

-- 
Richard Hobbs (Systems Administrator)
Toshiba Research Europe Ltd. - Speech Technology Group
Web: http://www.toshiba-europe.com/research/
Email: richard.hobbs at crl.toshiba.co.uk
Tel: +44 1223 376964        Mobile: +44 7811 803377



_____________________________________________________________________
This e-mail has been scanned for viruses by MCI's Internet Managed Scanning Services - powered by MessageLabs. For further information visit http://www.mci.com