Re: [nfsv4] Asking for details about OPEN_DOWNGRADE

[Rick Macklem <rmacklem at uoguelph.ca>]

On Wed, 25 Jun 2008, Trond Myklebust wrote:

> No. He is saying that it may make sense for a POSIX machine to add a
> DENY_WRITE while a file is being executed in order to enforce the
> ETXTBSY error that most people would expect if another process attempts
> an open for write.
> If you don't do an open_downgrade once you are done writing the
> executable then the above scheme would fail.

I can see why you might choose to do that and then, of course, open 
downgrades would be appropriate. (You could limit your implementation
to doing the downgrade for that specific case.)

> The other user of DENY_* would be Microsoft machines. If you don't plan
> to ever have your client run against a Windows-based server or work with
> other Windows-based clients, then you are fine, but if not, you are
> likely to be preventing the correct operation of many of their
> applications.
Lost me here. I can see why a WIN32 API client would do DENY_*, but I
don't see why having a server implemented under windows makes any
difference. Ditto w.r.t. other Windows clients. If your POSIX client
does DENY_WRITE, then another POSIX client won't be able to open the
file for writing and that will confuse it. Ditto for POSIX clients w.r.t
DENY_READ.

rick

_______________________________________________
nfsv4 mailing list
nfsv4 at ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4