LXD

LXD (Linux Container Daemon) is a Linux-based container virtualization and virtual machine management system. LXD is used to run several isolated Linux instances on a single node. LXD can be used as an alternative to OpenVZ technology.

Advantages of using LXD:

• container resources can be edited without stopping the container;
• high security for all cluster nodes;
• flexible configuration of cluster nodes;
• fast container creation and operation;
• high virtualization density.

Features of LXD clusters in VMmanager

In the current version of VMmanager, an LXD cluster can be created only with "Switching" or IP fabric network configuration type and ZFS storage. Containers and VM images are stored in the ZFS pool, and VM backups are stored on the cluster node. Ubuntu 20.04 OS must be installed on the cluster nodes.

The following is not available in LXD clusters:

• connecting user ISO images to virtual machines (VMs);
• adding nodes with two network interfaces;
• creating network bridges;
• live VM migration;
• installing Windows OS;
• connecting to the VM via VNC or SPICE.

VMmanager uses separate OS images to install operating systems in an LXD cluster. The following OS images are available in the current version:

For remote VMs administration you can go to the LXD container console. The console window opens in a separate browser tab.

Fine tuning of VMs

VMs in an LXD cluster have specific features in fine tuning:

• CPU and I/O weights can be set within the range from 0 to 10;
• Read and write speed can be set either in IOPS or Mbps;
• CPU emulation mode is not available;
• you can configure:
• • network weight — the bigger the weight, the more CPU time is allowed for request processing;
  • maximum number of active processes in the container;
  • CPU usage limit in percent;
  • creating containers inside a container using Docker, Podman, LXC, etc.

  Preparing the cluster node

  The LXD cluster node must have two disks: the first is the system disk, and the second is for creating VMs.

  Before connecting the server to the cluster:

  1. Configure the ZFS pool on the server:

     1. Install the zfsutils-linux utility:

        sudo apt install zfsutils-linux

     2. Define the block device of the second disk: 

        lsblk

        Example output

        NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
        loop0    7:0    0  91.9M  1 loop /snap/lxd/32662
        loop1    7:1    0  40.9M  1 loop /snap/snapd/20290
        loop2    7:2    0  49.3M  1 loop /snap/snapd/24792
        loop3    7:3    0  63.3M  1 loop /snap/core20/1828
        loop5    7:5    0  73.9M  1 loop /snap/core22/2045
        loop6    7:6    0 114.4M  1 loop /snap/lxd/33110
        sr0     11:0    1  1024M  0 rom 
        vda    252:0    0    50G  0 disk
        ├─vda1 252:1    0     1M  0 part
        └─vda2 252:2    0    50G  0 part /
        vdb    252:16   0    50G  0 disk

        In the example, the name of the second disk is vdb, so the name of the block device is /dev/vdb.
     3. Define the pool name. To do this, in the platform interface, go to Clusters → select a cluster → Parameters button → Local storages section → ZFS-pool parameter.
        
        

     4. Create the ZFS pool:

        zpool create <zpool_name> <device>

        Comments to the command:
        • <zpool_name> — pool name;
        • <device> — partition name.
     5. Add the pool to LXD: 

        lxc storage create <zpool_name> zfs source=<zpool_name>

        Comments to the command:
        
        • <zpool_name> — pool name.
  2. In the /etc/ssh/sshd_config configuration file specify the parameter: 

     UsePAM yes

     
     

  Useful tips

  Related topics:

  • Moving an OVZ container to LXD cluster