LXC Container show Load and hardware from Host

[nixmomo]

Hi,

i installed Proxmox 4 and in alle LXC Containers shows top and htop the hardware and load from my Host System and not from Container.

Is there anyway to change this that i only see in containers the hardware (CPU and RAM) thats configured in Proxmox Panel for this COntainer?

 

[nixmomo]

Has no one the same problem? :O

 

[willdashwood]

Yes I'm noticing this problem too. I was concerned that the load on newly deployed LXC containers was much higher than I expected. Then I realised load levels on the containers match exactly the host servers'.

 

[fabian]

You should only see the RAM and swap as configured, but the CPU related information is not hidden from an LXC container (except for uptime, which is calculated based on the age of the containers init process).

 

[alex3137]

I experience the same problem, a container shows all the cores when running htop.

 

[fabian]

This is an inherent limitation of lxc/cgroups/the linux kernel. See https://bugzilla.proxmox.com/show_bug.cgi?id=839 for an in-depth discussion.

 

[Spazmic]

There is no advantage in being able to see the host info in a container. It's a bug that is not fixable, and that was simply ignored due to the urgency of leaving openvz.

What a shame!

 

[gkovacs]

This is an inherent limitation of lxc/cgroups/the linux kernel. See https://bugzilla.proxmox.com/show_bug.cgi?id=839 for an in-depth discussion.

Actually, LXD + LXC 2.0 has some solutions for this problem, now the question is can they be implemented in Proxmox? More in this thread:
https://forum.proxmox.com/threads/revisiting-cpu-limit-for-lxc-2-0-lxd.28736/

I have also commented the relevant link in the bugzilla.

 

[Rhinox]

I have heard from many sides "LXC is not production-ready yet". Seems to be true (despite of LXC-devs thinking otherwise)...

 

[Plexus]

...also waiting for a solution for this...

 

[dcsapak]

there is a new options 'cores' for container which does something similar to lxd (pins container cores to real ones)
then the container should show only those cores in top/htop/etc.

 

[nixmomo]

In which Version?
I give a vm one core and see all cores from host and i see all host hardware in the proc

 

[fabian]

pve-container 1.0-79 added the option, 1.0-80 a new command to see the assigned cores
pve-manager 4.3-8 added the actual load balancing for containers with the cores option, 4.3-9 and 4.3-10 improved it further

 

[zima]

Is it numa aware when pining/repining cpus or just calculating cpus load and repining to free cpu? Can we have old "cpu limit" without pining to choose from in config for specific container? i don't mind container showing all cpus.

edit:Ii just found cpu limit is still available in options - not when creating container. so setting cores to total phisical cores count and using cpu limit i have old behaviour.

edit2:
after doing that cpu stats in gui are useless. setting cores=64 and cpu limit=2 we see cpu at 3.2% (100% is 64 cpus) when container use 100% of (2) cpu

 

[dcsapak]

you can delete the cores setting (in the gui just delete the number and ok; on the cli with pct set ID -delete cores)
it should behave like before then

 

[Sergey Semenov]

5.1 proxmox. And in all lxc-containers I see load average from proxmox host...

It is very bad. If I disable monitoring this parameter for all lxc-containers (by zabbix) - how we will known about problems in virtual-containers?

pve-manager/5.1-43/bdb08029 (running kernel: 4.13.13-5-pve)
root@vs04:~# pveversion -v
proxmox-ve: 5.1-38 (running kernel: 4.13.13-5-pve)
pve-manager: 5.1-43 (running version: 5.1-43/bdb08029)
pve-kernel-4.13.4-1-pve: 4.13.4-26
pve-kernel-4.13.13-5-pve: 4.13.13-38
libpve-http-server-perl: 2.0-8
lvm2: 2.02.168-pve6
corosync: 2.4.2-pve3
libqb0: 1.0.1-1
pve-cluster: 5.0-19
qemu-server: 5.0-20
pve-firmware: 2.0-3
libpve-common-perl: 5.0-25
libpve-guest-common-perl: 2.0-14
libpve-access-control: 5.0-7
libpve-storage-perl: 5.0-17
pve-libspice-server1: 0.12.8-3
vncterm: 1.5-3
pve-docs: 5.1-16
pve-qemu-kvm: 2.9.1-6
pve-container: 2.0-18
pve-firewall: 3.0-5
pve-ha-manager: 2.0-4
ksm-control-daemon: 1.2-2
glusterfs-client: 3.8.8-1
lxc-pve: 2.1.1-2
lxcfs: 2.0.8-1
criu: 2.11.1-1~bpo90
novnc-pve: 0.6-4
smartmontools: 6.5+svn4324-1
zfsutils-linux: 0.7.4-pve2~bpo9
ceph: 12.2.2-pve1

 

[Bidi]

Hello guys, i have the same problem on proxmox 5.2 is there a way to fix this ? all ower customers start to complay, why is my server fresh clean install and i have load of 3% :| is not ok at all

 

[yswery]

Was there ever a solution for this (I am on 5.3)

 

[Bidi]

Was there ever a solution for this (I am on 5.3)

I dont think so and is truly bad, i`m thinking of migrate back to 3...., since lxc i got lot of problems with this load

 

[roberto_fx]

I'm on 7.3 and can see the same issue (5 years later). This is very annoying as it becomes difficult to troubleshoot the server. I thought cgroups handles this Is this an issue with lxc in general?