The ability to connect between nodes, avoid mandatory restart, avoid connection limits and keep websocket connections alive easily gives compute engine a significant advantage over platforms like Heroku.
As far as I understand Heroku lets you connect between nodes (Dynos within a private space with dns discovery [0]) and also supports websockets [1]... for the mandatory restart, wouldn't you want to design your process as stateless anyway?
[0] https://devcenter.heroku.com/articles/dyno-dns-service-disco... [1] https://devcenter.heroku.com/articles/websockets
Sorry if I'm missing something obvious!
Maybe you can convince the powers that be to do something about these. :)
With most things you’d want stateless. Erlang/Elixir is designed with 3 built in databases, the ability to hold states on isolated processes and per process supervision and recovery. It’s a different animal than most.
It is wonderful that Heroku now offers exec, which at least allows using observer and other remote debugging tools. The tooling keeps getting better—but there is no denying that it was designed for Ruby deployment, and working the BEAM has significant production differences.
That's exactly it. The value proposition for Erlang & Elixir is essentially that they provide great abstractions for managing state and they make great back-end languages for creating stateful applications.