3 ms·

As an SRE who has had to baby sit production elixir apps I'd highly recommend first having a talk to your operations guys. I've never had a more challenging sys

As an SRE who has had to baby sit production elixir apps I'd highly recommend first having a talk to your operations guys. I've never had a more challenging system to debug and keep running. It's only a matter of time until you hit something as confusing as https://elixirforum.com/t/dbconnection-connectionerror-pid-xxx-exited-for-some-sql-queries/27478/4 https://elixirforum.com/t/dbconnection-connectionerror-pid-x...

devoutsalsa 5y ago
Yeah, I concur. If you’re a ops infrastructure let’s people deploy whatever ever they want in a container, and you can make the devs debug the elixir specific code, then you should be ok. If t try us devs force having to understand Erlang releases, GenServer timeouts, and Ecto based macro errors onto the ops group, everyone is gonna have a bad time.
1_player 5y ago
> I've never had a more challenging system to debug and keep running. Heh, the core system I put in place at work 3 years ago and is updated daily has never gone down once. Isn't that the point of the BEAM and supervision trees? It's different, requires to learn a new skillset, but I wouldn't call it "challenging to keep running."
SZJX 5y ago
Haha I'm not sure if we know each other since I know the issue you describe in this link. Regardless, I think this is a somewhat weird example since the it should not be DevOps' duty to look at and try to understand an error that stems from Ecto. And our company did face this issue, but this was largely related to the DB optimization and the Postgres instance instead of anything language-specific. If you have bad DB schemas and optimizations you're gonna face the same issue no matter in Java or Go. Saying that you've "never had a more challenging system to debug and keep running" sounds really dramatic as our Elixir apps have been running fine 99% of the time and barely even triggered any on-call incidents.

devoutsalsa 5y ago

Yeah, I concur. If you’re a ops infrastructure let’s people deploy whatever ever they want in a container, and you can make the devs debug the elixir specific code, then you should be ok. If t try us devs force having to understand Erlang releases, GenServer timeouts, and Ecto based macro errors onto the ops group, everyone is gonna have a bad time.

1_player 5y ago

> I've never had a more challenging system to debug and keep running. Heh, the core system I put in place at work 3 years ago and is updated daily has never gone down once. Isn't that the point of the BEAM and supervision trees? It's different, requires to learn a new skillset, but I wouldn't call it "challenging to keep running."

SZJX 5y ago

Haha I'm not sure if we know each other since I know the issue you describe in this link. Regardless, I think this is a somewhat weird example since the it should not be DevOps' duty to look at and try to understand an error that stems from Ecto. And our company did face this issue, but this was largely related to the DB optimization and the Postgres instance instead of anything language-specific. If you have bad DB schemas and optimizations you're gonna face the same issue no matter in Java or Go. Saying that you've "never had a more challenging system to debug and keep running" sounds really dramatic as our Elixir apps have been running fine 99% of the time and barely even triggered any on-call incidents.