this has been brewing in my head some time and it came up again while reading Adopting Elixir.
GenServers, supervisors etc. are great technologies that help you solve problems. They’re one of the things that is most special about elixir/Erlang. As a result lots of conference talks, blog posts etc. focus on them and it seems everyone wants to use them.
However, do you need them all the time? At least while using a framework (like Phoenix), chances are you don’t. Of course, until you got a problem that these help you solve.
Building a relatively standard CRUD web application with Phoenix? No need.
Just using channels for a chat like applications in Phoenix? You’re good.
The hidden detail of course is that you are using GenServers and friends without even knowing it - Phoenix runs every request and every channel in their own processes. Ecto uses poolboy for your database connections. It’s already parallelized and you don’t need to take care of it. That’s the beauty of it. What I’m saying is that in the standard situation the eco system takes care of you.
Why am I picking up this topic?
It feels like we talk so much about GenServers etc. that people who come to Elixir feel like they need to use them or they are not “really” using Elixir. I hear people say something to the tune of “We’re still using this like Rails - we should use GenServers” - without any need (granted they mostly don’t know what Phoenix & friends already do under the hood). At worst (as I’ve seen in some questions here) people create a single GenServer that then essentially all traffic needs to go through complicating their code while also adding an unneeded bottleneck. Maybe they just complicate their code, that’s also bad.
To get back to “Adopting Elixir” an example from it:
A new developer team started building their Phoenix applications.
They had always heard GenServers could be treated like microservices but even
tinier. This “wisdom” led them to push all of their database access control to
performance was abysmal. Under high-enough load, some pages took 3 sec-
onds to render because they built a bottleneck where none existed. They
defeated Ecto connection pools because all access happened through a single
In essence, they made it easy to create global, mutable variables in Elixir. They
essentially crippled the single biggest advantage of functional languages, for
no gain whatsoever.
Which is also what I’ve seen around a bunch of times. The book also provides some guidance as to what to best use GenServers for:
- Model state accessed by multiple processes.
- Run multiple tasks concurrently.
- Gracefully handle clean startup and exit concerns.
- Communicate between servers
So, what do I want in the end?
Well, I want to discuss with you all about this and hear your opinions!
I think we should make it clearer that you don’t have to use GenServers and that doing so might actually be harmful. My 2 production applications include no single GenServer written by us. They run fine. In general the eco system takes good care of you so you’re using them without realizing it (which is good imo).
I’m not saying you shouldn’t learn about GenServers. You should. But know when to use them and when not to.
Lastly, if you disagree I want you to scream at me and teach me the error in my ways