AlmaLinux 10.0

Option A: Azure Portal

• Open the Azure Marketplace and search for "AlmaLinux 10.0 SecureAnyCloud"
• Select the image and click Create
• On the basics tab, set:
• Subscription: Select your Azure subscription
• Resource Group: Create new or select existing
• Virtual Machine Name: Enter a name for your VM
• Region: Select your preferred Azure region
• Size: Standard_B2s recommended
• Under Administrator Account, select SSH public key and enter your key
• Under Inbound Port Rules, allow SSH (port 22)
• Click Review + Create, then Create

Option B: Azure CLI

az vm image terms accept --urn sac1647283583153:almalinux-10-0:default:latest
az vm create \
  --resource-group myResourceGroup \
  --name my-almalinux-10-0-vm \
  --image sac1647283583153:almalinux-10-0:default:latest \
  --size Standard_B2s \
  --admin-username azureuser \
  --generate-ssh-keys \
  --public-ip-sku Standard

Run these on your own workstation, not on the VM. Find the public IP, then connect:

az vm show --resource-group myResourceGroup --name my-almalinux-10-0-vm --show-details --query publicIps -o tsv ssh azureuser@<vm-ip>

There is no password to enter. If SSH asks you for one, the key you supplied at create time is not the key your client is offering.

Check that the machine is the AlmaLinux 10.0 release you expect:

cat /etc/almalinux-release grep -E '^(NAME|VERSION|ID|PLATFORM_ID)=' /etc/os-release uname -rm

Expected output:

AlmaLinux release 10.0 (Purple Lion) NAME="AlmaLinux" VERSION="10.0 (Purple Lion)" ID="almalinux" PLATFORM_ID="platform:el10" 6.12.0-55.43.1.el10_0.x86_64 x86_64

The kernel version carries an el10_0 tag. That tag marks the machine as being on the 10.0 stream. Because this image is frozen to the 10.0 vault, the tag stays el10_0 for the life of the machine.

The image is fully synced to the 10.0 vault at build time. dnf check-update returns exit code 100 when updates are pending against the enabled repositories and 0 when none are. Because the machine is pinned to the frozen 10.0 vault, it reports 0:

sudo dnf check-update echo "check-update exit code: $?"

Confirm no reboot is outstanding. needs-restarting -r returns 0 when nothing requires a restart:

sudo dnf needs-restarting -r echo "needs-restarting exit code: $?"

List the enabled repositories to confirm the update path is the vault:

dnf repolist --enabled

Expected output:

repo id repo name appstream-vault-10-0 AlmaLinux 10.0 - AppStream (vault, frozen) baseos-vault-10-0 AlmaLinux 10.0 - BaseOS (vault, frozen) crb-vault-10-0 AlmaLinux 10.0 - CRB (vault, frozen) extras-vault-10-0 AlmaLinux 10.0 - extras (vault, frozen)

Package signatures are verified on every install: no repository on this image sets gpgcheck=0, and the vault packages are signed with the same AlmaLinux 10 GPG key.

The frozen vault is the defining property of this image and the reason to choose it over the rolling AlmaLinux 10 image. Every repository points at the archived 10.0 vault at repo.almalinux.org/vault/10.0/. The rolling mirrorlist and metalink resolvers are disabled, so nothing can pull the machine forward to a newer minor:

grep -h baseurl /etc/yum.repos.d/almalinux-vault-10.0.repo | head -2
grep -rlE '^[[:space:]]*(mirrorlist|metalink)=' /etc/yum.repos.d/ 2>/dev/null || echo "no rolling resolver present"

Expected output:

baseurl=https://repo.almalinux.org/vault/10.0/BaseOS/$basearch/os/ baseurl=https://repo.almalinux.org/vault/10.0/AppStream/$basearch/os/ no rolling resolver present

Unattended updates are armed, but against the frozen vault. dnf-automatic applies any errata published within the 10.0 stream and never rolls the machine to a newer minor:

systemctl is-enabled dnf-automatic.timer grep -E '^(upgrade_type|download_updates|apply_updates)' /etc/dnf/automatic.conf

Expected output:

enabled upgrade_type = default download_updates = yes apply_updates = yes

To track whichever AlmaLinux 10 release is current, use the SecureAnyCloud AlmaLinux 10 image instead, which follows the rolling stream. Alternatively, convert this machine to the rolling channel yourself by restoring the stock repositories:

echo "to move to the rolling 10 stream: sudo rm /etc/yum.repos.d/almalinux-vault-10.0.repo && sudo dnf config-manager --set-enabled baseos appstream extras && sudo dnf distro-sync -y"

That reverses the pin and lets the machine move to the current 10 series minor. Leave the vault in place to keep the machine reproducible.

Confirm that SELinux is enforcing, root is locked and SSH is key-only:

getenforce grep -E '^[[:space:]]*SELINUX=' /etc/selinux/config
sudo passwd -S root
sudo sshd -T | grep -E '^(permitrootlogin|passwordauthentication|pubkeyauthentication|kbdinteractiveauthentication)'

Expected output:

Enforcing SELINUX=enforcing root L 2009-12-22 -1 -1 -1 -1 permitrootlogin without-password pubkeyauthentication yes passwordauthentication no kbdinteractiveauthentication no

without-password is how OpenSSH reports the prohibit-password setting. Both names mean the same thing: root may not authenticate with a password. passwd -S root reporting L means the root password is locked.

The commands check both the runtime mode and the persisted /etc/selinux/config value, because a runtime setting that is not persisted silently reverts on the next reboot.

Confirm the Azure Linux Agent, cloud-init and time synchronisation are healthy, and check your resources:

systemctl is-active waagent
systemctl is-enabled waagent cloud-init --version
systemctl is-active chronyd df -h / free -h

waagent is what lets Azure provision your SSH key, resize the OS disk, run extensions and report VM health, so it should always be active and enabled.

Note that free -h reports 0B of swap. This is deliberate. Azure manages swap on the ephemeral resource disk through waagent rather than on the OS disk, and a swap file baked into a Marketplace image fails certification. If your workload needs swap, configure it through waagent.

The Azure Network Security Group controls inbound and outbound traffic. It is the first, and usually the only, place you need to open a port.

Confirm what is listening before you open anything:

ss -tlnu

On a fresh machine SSH on port 22 is the only externally bound listener. Chrony appears on 127.0.0.1:323, which is loopback only and not reachable from the network.

The Cockpit web console is disabled on this image. An operating system base should expose exactly one service, so this image disables the socket and keeps the port closed, while leaving the Cockpit packages installed so you can turn it on in one command:

echo "cockpit.socket: $(systemctl is-enabled cockpit.socket 2>/dev/null || echo disabled)"
echo "to enable the Cockpit web console on 9090: sudo systemctl enable --now cockpit.socket"

rpcbind is disabled on this image. The stock cloud image enables it because nfs-utils depends on it, which leaves TCP and UDP port 111 open on every interface. NFSv4, the default in AlmaLinux 10, does not use rpcbind at all, so this image disables it and closes the port while keeping nfs-utils installed and fully working for NFSv4 mounts:

echo "rpcbind.socket: $(systemctl is-enabled rpcbind.socket 2>/dev/null || echo disabled)" rpm -q nfs-utils

If you specifically need NFSv3, which does require the portmapper, turn it back on with one command:

echo "to re-enable NFSv3 support: sudo systemctl enable --now rpcbind.socket"

Both services are disabled rather than masked precisely so those commands work without further steps.

If your policy requires a host firewall as well, install and enable it yourself:

sudo dnf install -y firewalld
sudo systemctl enable --now firewalld
sudo firewall-cmd --permanent --add-port=443/tcp sudo firewall-cmd --reload

Once firewalld is running, a port must be open in both the Azure NSG and firewalld before traffic reaches your service.

AppStream carries the application packages, served from the 10.0 vault. Search it, then install what you need:

dnf search nginx

Install a package and enable it in the usual way, for example sudo dnf install -y nginx followed by sudo systemctl enable --now nginx. Everything you install comes from the frozen 10.0 vault, so a package you add today resolves to the same version on a machine you build from this image next year.

To add another administrator, create the account with sudo adduser <name>. Add it to the wheel group with sudo usermod -aG wheel <name>. Then copy your public key into /home/<name>/.ssh/authorized_keys with mode 600, and set the .ssh directory to mode 700, owned by that user. The image ships with no human accounts, so azureuser is the only login until you add one:

getent group wheel