Libraries can start processes too

https://mattmower.com/2021/08/01/libraries-can-start-processes-too/

by sandbags • 5 years ago

70 30 5 years ago

dceddia• 5 years ago

In earlier versions of Elixir where you had to explicitly add the libraries as additional processes to start, I remember being very confused why things like e.g. an HTTP client needed its own process. To be honest I still felt a little uneasy about it every time. Thinking of those processes as just “the library needs some internal state” makes so much more sense!

dnautics4• 5 years ago

Problem is, it's not entirely clear that an http client library needs external state. (And for example, mint does not).

masklinn2• 5 years ago

Persistent connections? Cookies? Response cache? Conditional requests?

AlchemistCamp• 5 years ago

Many, many libraries do this! It's a key feature.

(to clarify, these are light-weight Elixir processes, not OS processes)

jolux3• 5 years ago

This is a huge part of the power of Elixir/Erlang. Libraries can be written to use the same powers of isolation, resiliency, and concurrency that application code has, at no extra cost to the consumer and with total transparency.

You quickly realize that every function you call could be working asynchronously underneath, and it’s not scary because the whole system is designed to function like this.

dnautics4• 5 years ago

I really wish that in elixir we drew a distinction between dependencies that do and don't spawn their own supervision trees. It would be nice if this were reflected in, say, hex.pm, possibly using different tags for them.

lostcolony• 5 years ago

It's not obvious from use? It's been a long time since I was in this world (well, Erlang), but it was application:ensure_started(App) that let me know "this dependency has its own supervision tree".

dnautics4• 5 years ago

Httpoison, for example, starts its own supervision tree to manage client process pools (not obvious that an http client should do that, definitely usecases where it shouldn't) and there is no indication either in mix.exs, or `MyApplication.Application.start/2` that Httpoison needs the tree. For most deployments, it's probably not a big deal. Some process will try to read http content, fail if httpoison isn't quite ready, and be restarted by its supervisor.

However, if you try to use it early in compilation or test, say, in your test_helper.exs file, speaking from experience, you could wind up with a very difficult to understand race condition where the httpoison process tree hasn't fully booted and you're trying to fetch something off the internet, and you don't have the same level of supervision protection -- if test_helper fails the whole test suite gives up and doesn't restart -- for obvious reason.

For the http case thankfully the elixir ecosystem is getting Mint as a base http library, which doesn't require a process tree out the gate, and several interesting explicit process-pool-libaries (finch, mojito) which are tuned for their own use cases that derive from mint.

QuinnWilton5• 5 years ago

Agreed. I also wish fewer libraries started their own supervision tree, and instead gave you a child spec to drop into your supervision tree. There's definitely use-cases where shipping libraries as an application makes sense, but oftentimes that sort of design causes problems for me, because it means not being able to start multiple copies of the dependency with different configurations.

I think Phoenix PubSub is a perfect example of how libraries should be structured, in that you just need to drop the module + options into your supervision tree, and you have the freedom of starting multiple independent copies of the tree, in different contexts, and with their own configurations: https://hexdocs.pm/phoenix_pubsub/Phoenix.PubSub.html#module...

jolux3• 5 years ago

You do sometimes have to be careful about how you handle configuration with embedding multiple copies of other supervision trees though: https://ninenines.eu/docs/en/ranch/2.0/guide/embedded/

dnautics4• 5 years ago

I'm 50-50 on that one (used to agree with you more but have since retraced a bit). This may be an overly nitpicky detail, but I you sort of want your own sup tree to not necessarily have a different-ly scoped "microservice" tied to it in terms of failure domains, and also just plain visual organization in your observer/livedashboard. For the 90% use case (e.g. http process pools) an indepentent sup tree is correct, but to your points,

1. it would be nice to have a choice. The library-writer should think about their users and choose which case is more correct. And make it opt-out and easy (let's say 2-3 loc) to implement the "other case", and spelled out explicitly in the readme/docs landing page.

2. PubSub indeed made (IMO) the correct choice when it migrated over from being its own sup tree to moving into the app's sup tree.

Thank you for listening to my TED talk.

derefr• 5 years ago

Alternately, the dependency can start its own supervision tree with any global processes/tables hanging off it from the beginning; and then export a Mod:start_link/1 function which clients will call, which will 1. start a child tree owned/managed by the dep's supervision tree; but which then 2. links that child subtree's root into the caller as well.

Such deps are integrated, by adding a stub GenServer that calls Mod:start_link/1 in its init/2 callback; and then adding a child-spec for that stub GenServer in your client app's supervision hierarchy.

The ssh daemon module in the stdlib works this way. Most connection-pooler abstractions (e.g. pg2, gproc) do as well.

sandbags3• 5 years ago

IIYC you're suggesting that what I am depending upon here is convenient but problematic?

My understand is not yet sophisticated enough to follow your point about "not being able to start mutiple copies of the dependency with different configurations".

Do you have any explanatory examples that could help me (and presumably others like me)? Thanks. m@t

Fire-Dragon-DoL2• 5 years ago

If a library is not written like that, it's poorly designed. It could provide one global started version as a commodity, but not being able to start it multiple times would be a big no.

Fire-Dragon-DoL2• 5 years ago

I think the manifest file already contains this information, so it's a matter of just surfacing this. At least for dependencies built using mix.

OskarS2• 5 years ago

I think Erlang and Elixir are incredible, and this particular feature of libraries starting BEAM processes seems cool in the abstract. However, this PARTICULAR task, Erlang/Elixir might be the worst possible language you could choose:

The Elixir approach to shared mutable state is wrapping it in a process. In this case, I needed a counter and the easiest way to implement it is to use an Agent which is a kind of process designed to handle simple state. In this case, the get_and_update function allows me to return the counter and increment it as an atomic operation.

This is literally just an atomic counter. It’s a single CPU instruction that is guaranteed to be safe. I don’t care how lightweight a BEAM process is, it’s not faster than updating an atomic counter. Doing it this way is also absurdly more complicated than using, say, a std::atomic in C++ (or the equivalent type in other languages).

Again, I think Elixir is cool, but if you want to show off how cool it is, maybe don’t use an example that is incredibly much slower and more complicated than it should be. It’s not a great look for Elixir.

enjo• 5 years ago

Of course you're right, but I think anyone who is capable of getting their head around Elixir is capable of understanding that this is a trivial example meant to showcase a powerful language feature. What you choose to use it for is up to you, not the article's author.

OskarS2• 5 years ago

I see that point, but it was pretty jarring reading this Elixir article as a curious outsider that they had to make these kinds of very advanced contortions to do something so simple in such a non-performant way. It’s something that is only impressive if you were already totally sold in on the BEAM way of doing things, but looks insane to any C/C++/Rust/Java/whatever developer.

whalesalad• 5 years ago

Processes on the BEAM may not be on the same machine or even in the same datacenter. You can't assume that everyone has access to the same physical memory.

sandbags3• 5 years ago

If I understand you correctly your comment boils down to whether or not you agree with the approach that Elixir and Erlang take to mutability.

In other langauges (even in Clojure, where I have a lot more experience) there are mutable global variables that you can use for these purposes. Unless I missed them then in Elixir/Erlang there are not.

Elixir/Erlangs approach to mutable state is the process. It doesn't matter what your specific application of shared mutable state happens to be.

We could discuss how any specific use case is more-or-less performant, but I would conjecture that someone who cared signficantly more about low-level performance over other things (e.g. systemic robustness) probably doesn't start with a BEAM language.

Libraries can start processes too

Related Stories