Host Machine Hangs After Ceph Setup on Nested Virtualization

Ripon

New Member
Feb 2, 2026
6
1
3
I am a full-time Linux user. I use **Zorin OS** on my base machine.
My base machine configuration is:

  • Intel Core i5 6th Gen (4 cores)
  • 16GB DDR RAM
  • 512GB SSD

On this system, I practice Proxmox VE using virtualization through Virtual Machine Manager (libvirtd).Below is the description of my lab setup:

pve1

  • storage-1 (20GB) + storage-2 (20GB) → Used for OS RAID1
  • storage-3 (20GB) + storage-4 (20GB) + storage-5 (20GB) → Used as 3 OSDs for Ceph
  • storage-6 (25GB) → Used as cache

pve2
  • storage-1 (20GB) + storage-2 (20GB) → Used for OS RAID1
  • storage-3 (20GB) + storage-4 (20GB) + storage-5 (20GB) → Used as 3 OSDs for Ceph
  • storage-6 (25GB) → Used as cache

pve3
  • storage-1 (20GB) + storage-2 (20GB) → Used for OS RAID1
  • storage-3 (20GB) + storage-4 (20GB) + storage-5 (20GB) → Used as 3 OSDs for Ceph
  • storage-6 (25GB) → Used as cache


In this setup, after running the PVE nodes for some time, my base machine becomes unresponsive (hangs), and nothing works unless I force a restart.However, this issue did not occur before configuring Ceph. Is this problem caused by Ceph or something else? What could be the solution?
 
Hi Ripon,
there can be multiple reasons für this kind of behaviour and the root cause can also be caused by the Hypervisor/ Host.

The described behaviour could probably point to a out of memory situation (especially with only 16GB RAM). You should be able to easily check this, by starting the VMs one by one while watching the memory consumption via
Code:
watch free -h

For more detailed troubleshooting it will be helpful, if you can provide related logs.

BR, Lucas
 
  • Like
Reactions: Ripon and UdoB
Hi Ripon,
there can be multiple reasons für this kind of behaviour and the root cause can also be caused by the Hypervisor/ Host.

The described behaviour could probably point to a out of memory situation (especially with only 16GB RAM). You should be able to easily check this, by starting the VMs one by one while watching the memory consumption via
Code:
watch free -h

For more detailed troubleshooting it will be helpful, if you can provide related logs.

BR, Lucas
I used htop to monitor my base machine and noticed that the system was using swap.Thank you brother
 
  • Like
Reactions: bl1mp