Todd <(E-Mail Removed)> wrote:
> I have a machine with a bonded NIC. I am trying to determine whether
> the bonding is doing anything for me. I have another machine, that
> is identical, that does not have a bonded NIC.
> I'd like to run a network throughput test on each of the machines
> that will show me how the network is performing. In other words, I
> want to know if un-bonding the NIC would hurt me or have no affect.
What bonding mode is being used?
> Both machines are being used as NFS servers. The bonded machine is
> MUCH more heavily used, but the non-bonded node will also be heavily
> used, eventually.
In broad handwaving terms, and with the exception of the "rr" bonding
mode, bonding will not allow a _single_ stream of traffic to run any
faster than before. It is the aggregate of the streams that is
increased via bonding. Are there _many_ clients accessing the NFS
server(s)? Are they on the same local LAN, or are they on the other
side of a router?
There's always netperf - the extent to which it simulates NFS traffic
depends on how you tell it to run.
http://www.netperf.org/
rick jones
--
firebug n, the idiot who tosses a lit cigarette out his car window
these opinions are mine, all mine; HP might not want them anyway...

feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...