🚀 go-pugleaf

> > scotta@jolt.mpx.com.au (Scott A Howard) writes:

> > So what is the *real* difference between reading a file 1 second before
> > it is deleted on a local filesystem, and reading a file 1 second after
> > it has been deleted on a remote filesystem, because it is cached?
> > 
> > Sure, you get data tha may not be quite current, but can happen at
> > anytime on a multiuser system.  And althogh I have not had a close
> > look at the kernel NFS src, I see to get some caching on my systems.

The problem is doing things like this and getting the correct answer:

tail bigfile.new                                  # loads the cache
rsh server "cd $CWD; sort bigfile >bigfile.new"   # Change file on remote system
tail bigfile.new                                  # reads the old data

POSIX says this will work.  It will not work on a cached NFS system.  The
problem is not really with the caching, as much as that NFS protocal which
is stateless and thus does not have any way to coordinate file access on
different machines.  This does have some advantages, as you can actually
reboot the NFS server, and the clients which have open files will simply
wait for it to come back up and everyone will be happy.

Jim