Lemmy locks up and uses a lot of ressources after about a week

Had this happen 2 times already and increasing the max memory seemed to fix it. Does lemmy need that much ressources or am I doing something wrong? I'm on the latest stable version

38 points · 7 comments · view on lemmy.world

7 Comments

chris@l.roofo.cc · 4 pts · 2y (3 replies)

Did you tune your postgres for your ram size?

Oha@feddit.de · 4 pts · 2y (2 replies)

I might havent done that. dont remember anymore

chris@l.roofo.cc · 4 pts · 2y (1 reply)

4gb is not a lot for a db. And it has to share it with lemmy and pictrs. It might not be enough with the standard config. Even then it might struggle if you subscribe to a lot of communities. You have to remember that every instance replicates everything the users are subscribed to. I am not sure though because my server hosts several applications and I have enough RAM.

kbotc@lemmy.world · 1 pts · 2y

Yea, there’s no where near enough RAM in that system. Add RAM and I’d bet everything fixes itself for a bit.

SeeJayEmm@lemmy.procrastinati.org · 3 pts · 2y

I'm running mine on a VPS with only 2.5 GB of ram and a few other smaller containers and have room.

I'd agree with the others about checking your postgres mem config.

archomrade@midwest.social · 1 pts · 2y

I'm a Linux/selfhosting novice, but from what I understand, Linux (and proxmox) uses intelligent resource management. It should adjust CPU and memory demands between tasks and VMs. My small stack of VMs running a media server and omv is basically always redlining with memory, occasionally with cpu (I don't have a dedicated GPU for encoding though)

The storage is a different problem, but it doesn't surprise me that Lemmy would use quite a bit of storage with active federation and multiple users.

syd@lemy.lol · 1 pts · 2y
[ removed ]
russjr08@bitforged.space · 1 pts · 2y

Are there any logs from the container(s) when it locks up?