Container Runtimes and WasmEdge benchmarking on Arm

For me this article is flawed. Hello Worlds application are not good metric foundation to make such a comparison.

Also it is no surprising that a container cannot beat a sandboxed programming platform that wasm is. I do not know if there is some mode in Java that could be comparable.

Yeah. Why bother with wasm when you can write the app from scratch in Java.

Boot up time also tends to be irrelevant since the true slowdown comes from initializing other dependencies anyways.

Wasm looks relevant for existing apps in C++ that you wish to run from a web browser.

Meaningful benchmarks are hard. Having said that, in this particular context it is a good metric to measure the latency for the simplest possible function invocation as that is important for serverless at the edge (and why you put those workers/functions as close as possible to the clients in the first place)

Has anybody ever seen large scale applications in real life? Or is WASM more like a middleware layer target? I'm genuinely interested in this as I'm yet to see a company using WASM in production.

> normal (debugging within the container is a problem)

That's a medium difficulty then. Kata isn't a great development environment.