Suppose that you have a situation where you need transparent highavailability NFS for something that is read-only and updated onlyinfrequently. Instead of going to the expense and bother of settingup real shared storage, it's tempting to try to implement this bysetting up a number of fileservers with local copies of the filesystem,synchronizing it from a master machine with rsync , and then using aslight variant of the basic HA NFS setup .

Unfortunately, the tempting easy way doesn't work; you can't

The NFS protocol uses something called a 'file handle' to identifywhat file a given operation applies to; this is sort of analogous tohow a traditional Unix system internally identifies files by theirinode number (well, their device plus their inode number).

In theory a NFS filehandle is opaque and an NFS server can use anyscheme it wants in order to uniquely identify files. In practice thereare a number of constraints on how a NFS server can form filehandles,based on preserving as many Unix semantics as possible

A while back I discovered a limitation in Linux's policy basedrouting, where you couldn't use the straightforward means to flexibly route outgoing traffic ona dual identity machine over differentinterfaces (for example, to force all SSH traffic to flow over one link,regardless of the destination). At the time I wrote:

To fix this situation up, you need to change the source IP address ofthe packets to fix them up. Unfortunately the only way I know of doingthis is to use source-