[openib-general] Detecting when an RDMA writer process disappears

[Mike Heffner]

Is there any method by which a receiving process that is polling in 
preregistered memory regions for data from a sender performing RDMA 
writes, can detect if the sender is killed? Say by a SIGKILL signal? The 
RC connection is setup using the RDMA CM and there do not appear to be 
any CM events created on the event channel, nor does there appear to be 
any async. events created. Occasionally I will get a CQE failure on the 
QP, depending on where the communication flow is, that I can use to mark 
the connection failed, but this happens only about 50% of the time.

An alternative solution would be periodically sending "keep-alives" and 
detecting the CQE failure, but I'd be interested to know if there are 
any other options that don't require sending keep-alives.


Thanks,

Mike

-- 

   Mike Heffner <mike.heffner at evergrid.com>
   EverGrid Software
   Blacksburg, VA USA

   Voice: (540) 443-3500 #603