Re: [OMPI users] Message reception not getting pipelined with TCP

Thanks George for the explanation, with the default eager size, the first message is received *after* the last message is sent, regardless the progress thread is used or not. an other way to put it is that MPI_Isend() (and probably MPI_Irecv() too) do not involve any progression, so i naively

Re: [OMPI users] Message reception not getting pipelined with TCP

On Thu, Jul 20, 2017 at 8:57 PM, Gilles Gouaillardet wrote: > Sam, > > > this example is using 8 MB size messages > > if you are fine with using more memory, and your application should not > generate too much unexpected messages, then you can bump the eager_limit > for

Re: [OMPI users] Message reception not getting pipelined with TCP

Gilles Gouaillardet, on ven. 21 juil. 2017 10:57:36 +0900, wrote: > if you are fine with using more memory, and your application should not > generate too much unexpected messages, then you can bump the eager_limit > for example > > mpirun --mca btl_tcp_eager_limit $((8*1024*1024+128)) ...

Re: [OMPI users] Message reception not getting pipelined with TCP

Hello, George Bosilca, on jeu. 20 juil. 2017 19:05:34 -0500, wrote: > Can you reproduce the same behavior after the first batch of messages ? Yes, putting a loop around the whole series of communications, event with a 1-second pause in between, gets the same behavior repeated. > Assuming the

Re: [OMPI users] Message reception not getting pipelined with TCP

Sam, this example is using 8 MB size messages if you are fine with using more memory, and your application should not generate too much unexpected messages, then you can bump the eager_limit for example mpirun --mca btl_tcp_eager_limit $((8*1024*1024+128)) ... worked for me George, in

Re: [OMPI users] Message reception not getting pipelined with TCP

Sam, Open MPI aggregates messages only when network constraints prevent the messages from being timely delivered. In this particular case I think that our delayed business card exchange and connection setup is delaying the delivery of the first batch of messages (and the BTL will aggregate them

[OMPI users] Message reception not getting pipelined with TCP

Hello, We are getting a strong performance issue, which is due to a missing pipelining behavior from OpenMPI when running over TCP. I have attached a test case. Basically what it does is if (myrank == 0) { for (i = 0; i < N; i++) MPI_Isend(...); } else { for (i =