(Sorry, I know nfs is boring.)
The linux nfs(5) manpage says"
NB: A so-called "soft" timeout can cause silent data
corruption in certain cases. As such, use the soft
option only when client responsiveness is more important
than data integrity. Using NFS over TCP or increasing
the value of the retrans option may mitigate some of the
risks of using the soft option.
Which mount_nfs(8) says nothing about. I don't see how a soft mount
could cause data corruption, unless you're using a buggy program that
doesn't handle error'd reads and writes properly, and this
<http://books.google.ca/books?id=AEV1x84DVkUC&pg=PA249&lpg=PA249&dq=nfs+soft+mount+data+corrupt&source=bl&ots=nDt9Rs8pYv&sig=28FGqCe9ialQVGU9QXvsiHhT1Ao&hl=en&ei=DHnnSa2BJIH0nQeB9eSIBw&sa=X&oi=book_result&ct=result&resnum=2>
seems to confirm that, but before I commit to using it can someone
tell me if there is a fundamental difference in Linux's implementation
or if they are just raising red flags and not remembering why?
Also, what is the difference between mount_nfs -i and mount_nfs -s?
Are the the same except that the first makes nfs return an error only
when I hit ctrl-c (or otherwise sent an interrupt) and the latter
makes nfs return an error whenever it times out? It seems like if you
want to kill a program you should always be able to kill it so I
suspect I'm misinterpreting -i.
Thanks!
-Nick
