[ofa-general] Re: Low NFS RDMA performance with Connect X

[Ramachandra K]

On 9/6/07, Michael S. Tsirkin <mst at dev.mellanox.co.il> wrote:
> > Quoting James Lentini <jlentini at netapp.com>:
> > Both the client and server code bases have been updated substantially
> > since the Mellanox SDK was released. Results are likely to be
> > different on the newer code.
> >
> > Finally, it is conceivable that there will need to be performance
> > tweeks for the Connect X hardware. For Tavor hardware, ULPs use a 1KB
> > MTU to achieve maximum performance (see the setup of the path_mtu QP
> > attribute in net/sunrpc/xprtrdma/verbs.c).
>
> One thing worth a try is interrupt coalescing.
> The simplest way to check is probably to apply the following patch
> and see if it helps. You can also try tweaking cq_max_count and
> cq_period module parameters.

I tried the interrupt coalescing patch, both with the default values
for cq_max_count and cq_period and also with some tweaking. Though I
could notice some small changes in numbers (both positive and negative
for different file sizes) these are still no where near the MT25208
numbers.

Regards,
Ram