That's all very well, but why *would* you use UDP if you can use TCP?
Sure, you need to overlap requests to get good behaviour out of either
NFS or UDP, but having done that, why go to all the effort of
reimplementing all of TCP's congestion control stuff?
> If by WebNFS, you mean nfs://server/path/to/some/file, then that's Brent's
> stuff from Sun and it has nothing to do with latency
No, it's a new revision of the protocol, which allows things like
looking up whole pathnames, rather than element at a time. It's Sun's
countermove to MS's proposal of CIFS as a standard filesystem protocol.
See http://www.sun.com/software/solaris/networking/webnfs/index.html
J
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu