Hi,

Do you know if and how it's possible to have the LinuxHA clustering setup use an OpenAFS share instead of the replicated block device it uses by default?

Basically, all nodes in the cluster, including balancers, are connected to the same OpenAFS network which contains two servers. When connected, these afs clients have access to one or more shared directories under /afs/[cellname]/[sharename]. Also the web server nodes have a local cache identical or rather well overshooting the maximum size of the data they are required to handle and serve to speedup service.

This may seem like a weird solution. But AFS would be required for a number of reasons, most of which is a kind of global "collaboration solution", for lack of a better term... And I really don't want to implement another shared storage solution when I already have one ready.

Since I'm only getting to know LinuxHA, can someone please help me understand how to make LinuxHA use the AFS share on each machine instead of the distributed block device?

Thank you sincerely!