Alan Cox wrote:
>
> > In summary: filehandles are useful at the protocol level, but it's
> > good for clients to retain pathnames if they need to recover
> > filehandles.
>
> How do you know the recovered pathname is the same file ?
>
> Alan
NFS v4 has an OPEN procedure, and servers can indicate to the client
if the file handle will be preserved through the duration of
an OPEN, including recovery through a server reboot. OPEN is true
state; NFSv4 is stateful, so servers that preserve file handles
during opens obviate the need for path name recovery.
The path name cache is there when the server won't promise to preserve
the file handle through OPEN, and even then, is best confined to
read-only file systems, or read-write file systems where the client
has a reasonable assumption that only it is messing with the file set
(e.g. I export a floppy drive on my lap top to my desk top
machine).
Eternal file handles are a (imho, necessary) artifact of a stateless design. With stateful design being practical in the age of rock solid servers, thin net free ethernets, and a lot more understanding of how to
solve the problem, eternal file handles are no longer a requirement, post
NFSv3.
-mre
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue May 23 2000 - 21:00:11 EST