gregory james marsh wrote:
Hi Carl,
A question from our group about the following client params you mentioned
in your email (included below):
--max-frame-size N (65535) the maximum frame size to request.
--bounds-multiplier N (2) bound size of write queue (as a
multiple of
the max frame size).
--tcp-nodelay Turn on tcp-nodelay
Could you further explain/clarify the behavior of "--bounds-multiplier".
We read over the code comments in
trunk/qpid/cpp/src/qpid/client/ConnectionSettings.h
It is used to determine how many messages/frames the producing thread in
the application can
get ahead of the IO layer. so 1 means lockstep, 2 is 2x etc...
A low value will decrease the time that a message can 'age' in the
buffer to be sent on the wire, but
could hurt throughput, a large number can increase throughput but hurt
latency (aging ).... The tuning
of this number obviously is network setup dependent.
Also, some basic server tuning can be done which depending on your gear
should improve your
numbers significantly.
With fame size, rough guess is to set it ~ 100bytes larger than the
message size you want to benchmark
for low latency.
regards
Carl.
