Deckhouse Virtualization Platform for bare metal

Prepare cluster nodes: configure the NFS server and worker node before platform installation.

Configure the NFS server

To configure NFS, complete the following steps:

1. Set up the NFS server for VM disk storage. Run the following commands on the NFS server:

   • For Ubuntu-based OS
   • For CentOS, Rocky Linux, ALT Linux, ROSA Server, RED OS, MOS OS

   sudo apt update
   sudo apt install nfs-kernel-server
   sudo mkdir -p <NFS_SHARE>
   sudo chown -R nobody:nogroup <NFS_SHARE>
   echo "<NFS_SHARE> <INTERNAL_NETWORK_CIDRS>(rw,sync,no_subtree_check,no_root_squash)" | sudo tee -a /etc/exports
   sudo exportfs -ra
   sudo systemctl restart nfs-kernel-server
   

   sudo dnf install -y nfs-utils
   sudo mkdir -p <NFS_SHARE>
   sudo chown -R nobody:nobody <NFS_SHARE>
   echo "<NFS_SHARE> <INTERNAL_NETWORK_CIDRS>(rw,sync,no_subtree_check,no_root_squash)" | sudo tee -a /etc/exports
   sudo exportfs -ra
   sudo systemctl enable --now nfs-server
   sudo systemctl restart nfs-server
   

2. Verify NFS access from the master node. Run the following commands on the master node:

   • For Ubuntu-based OS
   • For CentOS, Rocky Linux, ALT Linux, ROSA Server, RED OS, MOS OS

   sudo apt update
   sudo apt install nfs-common
   sudo mkdir -p /mnt/dvp-nfs-test
   sudo mount -t nfs4 <NFS_HOST>:<NFS_SHARE> /mnt/dvp-nfs-test
   sudo touch /mnt/dvp-nfs-test/probe && echo "OK: mount and write work"
   sudo rm -f /mnt/dvp-nfs-test/probe
   sudo umount /mnt/dvp-nfs-test
   sudo rmdir /mnt/dvp-nfs-test
   

   sudo dnf install -y nfs-utils
   sudo mkdir -p /mnt/dvp-nfs-test
   sudo mount -t nfs4 <NFS_HOST>:<NFS_SHARE> /mnt/dvp-nfs-test
   sudo touch /mnt/dvp-nfs-test/probe && echo "OK: mount and write work"
   sudo rm -f /mnt/dvp-nfs-test/probe
   sudo umount /mnt/dvp-nfs-test
   sudo rmdir /mnt/dvp-nfs-test
   

   The OK: mount and write work line means the node mounted the directory and wrote to it as root, so both networking and the no_root_squash option are in place. If the line is missing, use the error message:

   • mount.nfs4: Connection timed out — the NFS server is unreachable. Check the firewall and make sure its IP differs from the master and worker IPs.
   • mount.nfs4: access denied by server — the node is outside the subnet listed in /etc/exports on the NFS server.
   • touch: cannot touch ...: Permission denied — the directory is exported without the no_root_squash option.

Prepare the worker node

<CAPS_SSH_PUBLIC_KEY>

<CAPS_SSH_PRIVATE_KEY>

To continue setup, create the caps user by running the following commands on the worker node:

export KEY='<CAPS_SSH_PUBLIC_KEY>'
sudo useradd -m -s /bin/bash caps
sudo usermod -aG sudo caps
echo 'caps ALL=(ALL) NOPASSWD: ALL' | sudo EDITOR='tee -a' visudo
sudo mkdir -p /home/caps/.ssh
echo "$KEY" | sudo tee -a /home/caps/.ssh/authorized_keys
sudo chown -R caps:caps /home/caps
sudo chmod 700 /home/caps/.ssh
sudo chmod 600 /home/caps/.ssh/authorized_keys

export KEY='<CAPS_SSH_PUBLIC_KEY>'
sudo useradd -m -s /bin/bash caps
sudo usermod -aG wheel caps
echo 'caps ALL=(ALL) NOPASSWD: ALL' | sudo EDITOR='tee -a' visudo
sudo mkdir -p /home/caps/.ssh
echo "$KEY" | sudo tee -a /home/caps/.ssh/authorized_keys
sudo chown -R caps:caps /home/caps
sudo chmod 700 /home/caps/.ssh
sudo chmod 600 /home/caps/.ssh/authorized_keys

On Astra Linux with Parsec enabled, set the maximum integrity level for caps:

sudo pdpl-user -i 63 caps

The cluster nodes are ready for Deckhouse Virtualization Platform installation.