[mvapich-discuss] Verify the application is really running over IB

[Abhinav Vishnu]

Hi Henry,

To compile osu_latency.c with MVAPICH, you need to use to the MPI compiler,
mpicc, which must be shipped with topspin libraries. On the shell, please
do the following:

$ mpicc osu_latency.c -o lat

The same executable (lat) may be used for execution within a node and across
nodes. To verify that you are using IB, i would recommend using two 
nodes, each
having InfiniBand installation, like the Topspin. Say the nodes are 
node1 and node2.

MVAPICH provides mpirun_rsh to run the executable:

$ mpirun_rsh - np 2 node1 node2 ./lat

As Dr Panda mentioned in an earlier mail, if you are using Native 
InfiniBand, you should
see the latency in order of 2~4 us. Further options and instructions for 
execution are
present here:

http://mvapich.cse.ohio-state.edu/support/mvapich_user_guide.html#x1-200005.2

The MVAPICH user guide also has a detailed troubleshooting section, 
which is present
at the following URL:

http://mvapich.cse.ohio-state.edu/support/mvapich_user_guide.html#x1-300007

Please let us know the outcome of your experimentation.

Thanks and regards,

:- Abhinav
> hello, Dhabaleswar:
> I downloaded
> https://mvapich.cse.ohio-state.edu/svn/mpi/mvapich/trunk/osu_benchmarks/osu_latency.c
> how can I run the benchmark then? I mean how many cpus or nodes should be
> involved?
> Thanks a lot.
> Henry, Wu
> |
> |
> | On Wed, 29 Aug 2007 wgy at altair.com.cn wrote:
> |
> |> Hello, list:
> |> It might be a silly questions but I wonder how to verify run with
> |> mvapich
> |> (come with Topspin Infiniband) over Infiniband, NOT Gigabite network.
> |> Is there an option to force mvapich to use IB network otherwise just
> |> exits?
> |
> | MVAPICH has several underlying interfaces: Gen2, uDAPL, VAPI, TCP/IP and
> | shared memory. Please take a look at the user guide (available from
> | mvapich project page) to see the differences and capabilities of these
> | interfaces. Gen2 interface (corresponding to OFED) will give you
> | the best performance and scalability. If you have OFED stack installed,
> | you should be able to configure mvapich to run over Gen2 interface
> | (as per the instructions indicated in the user guide). During OFED
> | installation, you can also select mvapich from the package.
> |
> | On your existing installation, you can also run OSU benchmarks (such
> | as OSU latency). If you get latency number in the range of 2~4 microsec
> | for short messages (say 4 bytes), it is already running over the native
> | IB.
> |
> | Hope this helps.
> |
> | DK
> |
> |> Thanks for your suggestion.
> |> Rdgs.
> |> Henry, Wu
> |>
> |> _______________________________________________
> |> mvapich-discuss mailing list
> |> mvapich-discuss at cse.ohio-state.edu
> |> http://mail.cse.ohio-state.edu/mailman/listinfo/mvapich-discuss
> |>
> |
> |
>
>
> _______________________________________________
> mvapich-discuss mailing list
> mvapich-discuss at cse.ohio-state.edu
> http://mail.cse.ohio-state.edu/mailman/listinfo/mvapich-discuss
>
>
>