Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!uwm.edu!lll-winken!sun-barr!newstop!east!hinode!geoff
From: geoff@hinode.East.Sun.COM (Geoff Arnold @ Sun BOS - R.H. coast near the top)
Newsgroups: comp.protocols.nfs
Subject: Re: Interdrive or PC/NFS Connection Timeouts
Keywords: NFS Interdrive
Message-ID: <919@east.East.Sun.COM>
Date: 13 Oct 89 13:55:29 GMT
References: <6@softeng1.UUCP>
Sender: news@east.East.Sun.COM
Reply-To: geoff@hinode.East.Sun.COM (Geoff Arnold @ Sun BOS - R.H. coast near the top)
Organization: Sun Microsystems, Billerica MA
Lines: 29

In article <6@softeng1.UUCP> zumbachl@softeng1.UUCP (Lyle Zumbach) writes:
>
>Both Sun's PC/NFS and FTP's Interdrive don't seem to handle
>network connection loss very well. It seems like there is
>a high chance of losing data (and not knowing about it) if you
>lose your network connection. Is this true? Shouldn't the user
>be presented with the Abort, Retry, Ignore message (DOS critical
>error handler?) ? Does the fact that NFS is a stateless protocol
>cause the problem?

PC-NFS will indeed give you an "Abort, Retry, Ignore" on
timeouts. The number of retries is configurable (via the "/r"
switch); if you set "/r0" we'll retry forever (or until a
Ctrl-Break). The default number is 4. Note that we use an
exponential backoff up to a 30 second retry interval; the
initial value and the backoff rate depend on the network
connection.

As for losing data, the timeout condition is treated the same
as a Drive Not Ready on a diskette. If you select "Retry",
the system call is restarted by DOS, and  if the network
is now ok (e.g. the server has rebooted) you'll lose no data.

I'll let James comment on InterDrive.

Geoff Arnold,                              Internet: geoff@East.Sun.COM
PCDS Group, Sun Microsystems Inc.
---------------------------------------------------------------------------
"Who's next?" "Me, doctor?" "No, ME doctor, YOU patient." (Graham Chapman, RIP)