Roland Dreier wrote:
I found at least one silly bug in my code that could cause the crash
you saw. Can you pull the latest svn kernel code and try again?
By the way, I found regress.sh in the dapl source but I'm not sure
what it means to "set to 40x40." Can you explain that for a non-DAPL
expert?
Thanks,
Roland
sorry, dapltest has parameters for threads and connections per thread
that I use for my scale-up testing. I just took the "client 6" section
of regress.sh (see below) and changed to -t 40 -w 40 and ran against a
srv.sh running on a separate node. The device is the device name used in
/etc/dat.conf and the host is the hostname (IPoIB ip address) where
srv.sh is running.
The current version of uDAPL does not have the async processing hooks so
you may not see any problems. I am working on these changes now and will
send out a patch shortly.
I will give your latest changes a try.
#====================================================================
#client6
#====================================================================
./dapltest -T T -s ${host} -D ${device} -i 10000 -t 4 -w 8 \
client SR 256 \
server RW 4096 \
server SR 256 \
client SR 256 \
server RW 4096 \
server SR 256 \
client SR 4096 \
server SR 256
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
