Benchmark of resilience against Erlang

[joeblew999]

https://freedium.cfd/https://medium.com/@codeperfect/we-tested-7-languages-under-extreme-load-and-only-one-didnt-crash-it-wasn-t-what-we-expected-67f84c79dc34

erlang was the only one to lead the pack .

should do a resiliency benchmark with ergo perhaps 🤔??

[halturin]

There are no details about their benchmarks. What is the logic behind the processRequest(c) that led to leaks under pressure?

The Erlang "example" there is just the code of a supervisor. What does the connection_pool do?

No measurements, no numbers at all. It's a clickbait topic.

P.S. Take a look at the number of articles published by this author (for the last week) —seems like "vibe-writing."

[matheuscamarques]

I'm truly impressed with all the work you've done so far on this project and genuinely interested in your findings during this migration but let's get straight to the point.

A resilience test on a platform that was specifically designed to be resilient, and has been engineered, refined, and validated over 40 years of real-world use? Sorry, but what exactly were you expecting?

Resilience takes more than introspective optimism and wishful thinking, it’s the result of decades of engineering, hard lessons, and historical context.

The question you need to ask yourself now is whether you truly want a sample from this benchmark...

That said, I agree there's a lack of actual code and benchmark data.
However, disputing the historical facts behind this platform won’t be easy.

[halturin]

It is very naive to think that only Erlang can create truly reliable systems. If you're referencing 40 years of experience, then compare what Go has achieved in its 15 years. I would even say that in 2025, appealing to the myth of Erlang's super reliability against modern tools is the level of a novice.

[matheuscamarques]

In a few years you will converge on the same thought as mine. But I do not deny your efforts are interesting. But remember, BEAM does not stop, it is not stagnant in time.