Hi > What tools were you using for testing? What are the potential bottlenecks > with > your approach to the task at hand? Are the tools you're using suited to > the > tasks and approach you're using? > We are using JMeter and Apache Benchmark. JMeter is run on a client (my laptop) that connects to a server (AWS micro instance). Apache Benchmark was run on the server
Are you controlling for memory allocation? > No, we are not. But (this may be a stupid question), why should we have to (are we talking about memory leaks on Node??) > How much paralellism? How much allocation in the hot paths? How long lived > are the requests? > For Uwsgi (Python), we have 30 threads listening to incoming connections. > Can the database or database driver be serializing things unexpectedly? > I suppose it could. > What's your concurrency? What's the significance of the variation in > timing? > We don't really know the answer to the second part of the question (hence we are asking this). And how do we measure the first. > Are you testing locally or across a network? > Both. Jmeter accross a network, apache benchmark locally. > Did your micro instance get CPU throttled? Do you have metrics that would > let you know? > Perhaps this is happening, but it should still allow for Node to out perform Python as the majority of expensive operations were IO based. Still, I would expect Node to beat to Python, even if the exact results changed all the time, I would still expect Node to come out on tops. Things that you have mentioned above allow for variations in the results, but it does not explain why Node does not always beat Python. -- Job Board: http://jobs.nodejs.org/ Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines You received this message because you are subscribed to the Google Groups "nodejs" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/nodejs?hl=en?hl=en
