[Linux-cluster] NFS client failover

[Mike Cardwell]

crypto grid wrote:

>     I have a small cluster which shares a filesystem via GFS on a SAN,
>     with iSCSI. There are a couple of hosts external to this cluster
>     which I would like to have access to the GFS filesystem via NFS. I
>     have exported the same mountpoint on each of the hosts that have the
>     GFS mount.
> 
>     I can connect via NFS to one of these cluster hosts, but when it
>     goes down, I have to unmount it, and then remount against another
>     host. I suspect I could do something with a floating IP? but I was
>     wondering if it would be possible to do this with client side logic
>     so that the nfs mount automatically moves to a different working
>     host if one stops responding?
> 
> I don't see why you want to resolve this from the client side.

I figured that failover would happen more smoothly if the client was 
aware of and in control of what was going on. If the IP suddenly moves 
to another NFS server I don't know how the NFS client will cope with that.

> You must configure an ip address resource.

Yeah, that's what I was going to try if there wasn't a client side solution.

> Take a look at this document:
> http://www.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/pdf/Configuration_Example_-_NFS_Over_GFS.pdf

That document looks very useful. Thank you for pointing it out to me.

-- 
Mike Cardwell - IT Consultant and LAMP developer
Cardwell IT Ltd. (UK Reg'd Company #06920226) http://cardwellit.com/