VM’s crashing/stopped due to storage/memory limit?

[Boyd911]

Newbie question.
Am experimenting with 4 equal VM running Ubuntu. Each VM requires 8GB, 4 cores, 250 GB storage - my server has 12 cores, 32 GB (effectively 31.19GB), and 1TB (effectively 940GB). Running 3 VM’s —> CPU usage is low <30%, Memory usage is 85%, Disk usage is low.

Running 3 VM’s is no problem and they run smoothly. But as soon as I start VM 4, after a few minutes one of the running VM’s stops. What is causing this? Is it the lack of available memory or storage capacity?

Second question, are the VM required resources (cpu, mem, storage) reserved for that VM even if it’s not needing it? Or is Proxmox sharing the resources over all VM’s in case these aren’t used. trying to learn how this works - it feels cpu usage is share but memory and disks not.

 

[pos-sudo]

Hi There,

According to your first question; You have 4 VM's running with a configuration of 8GB RAM, 4 cores, 250GB storage each. While your node/host (server) has in total 12 cores (including hyperthreading?) 32GB of RAM installed and 1TB diskspace. The thing here is when you start spin VM's up with more resources then it could handle on de node then the OOM-killer would kill an VM in order to protect it from being overloaded the whole server. So I think it is good to be carefull when adding resources to VM's. Especially RAM and disk space.

An answer about your second question is according to RAM and CPU no, it is nog a "hard given" resource to that specific VM, however while booting a VM it will actually use 100% of its CPU and RAM. My advise is to configurate a VM with it specific needs so starting with 4GB RAM and maybe if this fits in your situation 100GB storage?

I hope this answer helps you further.