[plug] Anyone here interested in NFS and familiar with the kernel implementation?

[Ian Kent]

On Wed, 2006-09-20 at 12:10 +0800, Mark J Gaynor wrote:
> What does it entail?

Basically, NFS v3 and v2 failover.

For example userspace mount syntax like (there's other possible syntax):

mount -o ro s1,s2,s3:/dir/other /path

then NFS kernel client needs to identify when server is no longer
available and failover to another server.

There's code to create a list of servers ordered by ping time within
proximity (ie. network proximity) that would be passed to the kernel.
Initially we can use the simplifying assumptions that if locks are held
by the mounted fs we will not failover and the all the hosts in the
failover list must have the same NFS protocol version.

> 
> Mark
> --
> 
> *********** REPLY SEPARATOR  ***********
> 
> On 20/09/2006 at 11:26 AM Ian Kent wrote:
> 
> >Hi,
> >
> >I'm wondering whether anyone locally would be interested in working on a
> >little NFS project with me?
> >
> >Any takers?
> >
> >Ian
> >
> >
> >_______________________________________________
> >PLUG discussion list: plug at plug.org.au
> >http://www.plug.org.au/mailman/listinfo/plug
> >Committee e-mail: committee at plug.linux.org.au
> 
> 
> 
> _______________________________________________
> PLUG discussion list: plug at plug.org.au
> http://www.plug.org.au/mailman/listinfo/plug
> Committee e-mail: committee at plug.linux.org.au