[openib-general] IPoIB CM

Sean Hefty mshefty at ichips.intel.com
Wed Nov 29 10:48:05 PST 2006


Michael S. Tsirkin wrote:
> e. Some notes on code
> 1. SRQ is used for scalability to large cluster sizes
> 2. Only RC connections are used (UC does not support SRQ now)
> 3. Retry count is set to 0 since spec draft warns against retries
> 4. Each connection is used for data transfers in only 1 direction,
>    so each connection is either active(TX) or passive (RX).
>    2 sides that want to communicate create 2 connections.
> 5. Each active (TX) connection has a separate CQ for send completions -
>    this keeps the code simple without CQ resize and other tricks

Would you expect that the use of this code would likely result in creating 
all-to-all connections between all nodes?  In particular, I'm thinking about MPI 
applications that use all-to-all connections.  I realize this is done at the 
RDMA layer, but is there anything that would end up connecting all of the nodes 
via iopib?

- Sean




More information about the general mailing list