MariaDB is being killed due to OOM

hxii

I can't quite figure out why, but mariadb seems to be crashing/being killed every couple of days.
According to dmesg:
[3714356.085901] Out of memory: Kill process 17001 (mysqld) score 259 or sacrifice child [3714356.086327] Killed process 17001 (mysqld) total-vm:1004168kB, anon-rss:261008kB, file-rss:0kB, shmem-rss:0kB

Anybody ever run into this issue with Wordops?

portofacil

It seems your server has not enough RAM to run all services. Which hardware configuration are you using right now?

hxii

portofacil 1 vcore instance on Vultr. (1GB RAM, no swap)

portofacil

You need at least 2GB RAM and 1GB swap, according to my previous experience.

hxii

portofacil Thanks! This is weird however, as it usually happens when the instance is not under load + I can't recall this happening with EE.
In any case, I might look into upgrading if we cant pinpoint the cause.

JuanMaia

I use many VPS 1/1 vultr.
I always have this problem when everything is installed.
php72, php73, panel, manched, redis, phpmyadmin ... +, basically what WordOps has, then you will have an excessive use of ram, leaving little for the bank to work sometimes. Then he will be killed. Analyzing this happens a few times then I create the swap meme, if this happens several times in a day then a higher VPS is needed.