As @ttielebein introduced previously, one of the missions of the IOT Enterprise Deployment Center (EDC) is to publish benchmarks that showcase the ThingWorx Platform deployed to solve real-world IOT business problems.
Our goal is that these benchmarks can be used as a reference or baseline for architects working on their own implementations... showing not only a successful at-scale implementation, but also what happens when that same implementation is pushed to ...or even past... it's limits.
Please find the first installment attached - a reference benchmark demonstrating ThingWorx deployed to monitor 15,000 assets with a high-volume of data properties per asset. Over 250 hours of simulations were conducted as part of producing this benchmark.
The IOT EDC team will be monitoring this post (as well as our other posts in the IOT Tech Tips forum) to answer any questions we can about the approaches taken in designing, deploying and simulating this implementation.
As the team will publish more benchmarks like this will be published in the future, we also greatly value any feedback you have that can help us to improve the content for future documents.
Did you measure Connection Server nodes? Was it necessary to allocate so much for processing less than 200 messages per second overall?
Yes, you’re right to question the need for four connection servers - two connection servers would likely be sufficient for this deployment. It might even be stable with a single connection server, but best practice is to have at least two connection servers.
We used four connection servers to address some challenges with the simulators used for this simulation when we exercised the system with larger CPU configurations.
Thanks - Dave