Hi,

I came across the following bug. NFS writes from Redhat 7.3 client result in filer giving a lot of RPC timeouts for all other clients on the network. Here's a link describing what's known about this bug so far:

http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=65069

With default wsize value of 15K, tcpdump shows a lot of fragmentation packets.

While it's obvious that it's RedHat 7.3 bug, I am still wondering why NFS v3 is vulnerable to such sort of DoS attack.

On related note, I looked at registry file on the file, and I found these entries:

options.nfs.tcp.recvwindowsize=26280 options.nfs.tcp.xfersize=32768 options.nfs.udp.xfersize=32768

Is the 1st one only applicable to NFS over TCP? Is the last one read-only?

Thanks

Igor