NFS write corruption on 8.0-RELEASE

[Rick Macklem]

On Thu, 11 Feb 2010, John Baldwin wrote:

>>
>> Case1: single currupted block 3779CF88-3779FFFF (12408 bytes).
>> Data in block is shifted 68 bytes up, loosing first 68 bytes are
>> filling last 68 bytes with garbage. Interestingly, among that garbage
>> is my hostname.
>
> Is it the hostname of the server or the client?
>
Oh, I realized the first 4 bytes of the garbage is the record mark
that preceeds the RPC header for TCP, so the garbage is the first
part of the RPC after the TCP/IP header.

>
> Can you reproduce this using a non-FreeBSD server with a FreeBSD client or a
> non-FreeBSD client with a FreeBSD server?  That would narrow down the breakage
> to either the client or the server.
>
If using a non-FreeBSD client/server isn't convenient, another way would
be to do a binary packet capture (something like "tcpdump -s 0 -w <file> 
<host>") and then looking at it in wireshark for a failed case and see
if the data is corrupted on the wire.

rick