Gregory Maxwell wrote:
Because if the server goes away you get stuck unkillable processes.
NFS is great in the steady state, but during failure it tends to trigger bad and unexpected behaviour in everything around it.
Thinking about it, the problem is not the use of nfs, but that nfs driver performs quite badly on server unreachable conditions. How much would be needed to make them better? Nfs seems to already have several mount options for improving its behavior:
soft Soft-mount the driver. This option causes an I/O error to be reported to the process attempting a file operation when a major timeout occurs timeo=n This sets the time (in tenths of a second) the NFS client will wait for a request to complete. intr Allow signals to interrupt an NFS call. Useful for aborting when the server doesn't respond.