Is this technology being used anywhere commercially or with intent for commercial use? Can you also quantify how long/roughly how many thousands of hours of effort this has been in development?
Have you had a chance to measure operational throughput vs hardware utilization yet? I'd be very interested in loose metrics and stats instead of a general "faster than" statement.
Sorry for my slow response; how does the process handle scaling above 1kb? Traditionally there's a predictable curve of expected performance as the scale of request size increases.
Happy to take this conversation private; spent the last 10 years in performance and scalability as a indie consultant for Fortune companies. I know a lot of people looking for replacement mechanisms for high volume, variable sized transaction enterprise buses.
Also interested to understand the data access methods you've tested and what works best with your code. eengstrom >at< gmail.
Development started in late 2009 but there was a couple of "let's redo this part" and "ho, wait, it's never going to work!".