Getting Started

You’ll need a machine with KVM. If you can run ls /dev/kvm and see a file, you’re good.

What You Need

Hardware

Manager node can be small — it mostly coordinates things. Workers need more since they actually run VMs.

Node	Minimum	Works Better
Manager	1 vCPU, 1 GB	2 vCPU, 2 GB
Worker	2 vCPU, 4 GB	4 vCPU, 8 GB

Software

Ubuntu 20.04+ or Debian 11+ work well. Any KVM-compatible distro should do.

You need root access for setting up bridges and Firecracker.

Check KVM

ls -la /dev/kvm                    # Must show a file
lscpu | grep Virtualization        # VT-x (Intel) or AMD-V (AMD)

If you’re running inside a VM, nested virtualization has to be on:

cat /sys/module/kvm_intel/parameters/nested  # Should be 'Y'

# If it's 'N':
sudo modprobe kvm_intel nested=1

Install

One-Line Install (blessed path — ADR-005)

install.sh only downloads the latest release binary and verifies its checksum. Everything else is handled by the swarmcracker setup subcommand:

# 1. Install the binary
curl -fsSL https://raw.githubusercontent.com/restuhaqza/SwarmCracker/main/install.sh | sudo bash

# 2. Verify prerequisites (KVM, kernel modules, tools)
sudo swarmcracker setup check

# 3. Install Firecracker, jailer, kernel, and rootfs
sudo swarmcracker setup install --download-kernel --download-rootfs

# 4. Create the VM bridge + enable NAT
sudo swarmcracker setup network

# 5. Generate the config
sudo swarmcracker setup config --non-interactive

Repeat steps 2–5 on every node (manager and workers). Then start the cluster:

# On the manager node
sudo swarmcracker cluster init --advertise-addr <MANAGER_IP>:4242

# On each worker node
sudo swarmcracker cluster join --token <TOKEN> <MANAGER_IP>:4242

Build It Yourself

git clone https://github.com/restuhaqza/SwarmCracker
cd SwarmCracker
make all
sudo make install

The Kernel Thing

Firecracker needs an uncompressed ELF kernel at /usr/share/firecracker/vmlinux. swarmcracker setup install --download-kernel downloads a known-good kernel for you.

Don’t try downloading from GitHub raw URLs — you’ll get HTML, not a binary.

To extract one from your host kernel instead:

sudo mkdir -p /usr/share/firecracker
./test-automation/scripts/extract-vmlinux.sh /boot/vmlinuz-* /usr/share/firecracker/vmlinux

# Check it worked
file /usr/share/firecracker/vmlinux
# Should say: ELF 64-bit LSB executable, x86-64

Local Test Cluster

For a multi-node cluster with microVMs placed on separate nodes (VXLAN overlay), use the single-host lab:

sudo test-automation/multinode/cluster-lab.sh up 2   # create + provision + cluster
sudo test-automation/multinode/cluster-lab.sh test   # cross-host matrix
sudo test-automation/multinode/cluster-lab.sh destroy

Start the Cluster

Manager Node

sudo swarmcracker cluster init --advertise-addr <MANAGER_IP>:4242

--advertise-addr is critical: workers need to reach the manager, and without it they’ll try to connect to 0.0.0.0, which won’t work.

This starts:

SwarmKit manager (Raft consensus for cluster state)
Control socket at /var/run/swarmkit/swarm.sock
TLS certificates
Join tokens saved to /var/lib/swarmkit/join-tokens.txt

Get the Join Token

sudo swarmcracker cluster token worker

Look for the SWMTKN-... token in the output.

Join Workers

sudo swarmcracker cluster join <manager-ip>:4242 --token <TOKEN>

Check the Cluster

sudo swarmcracker node ls
sudo swarmcracker cluster health

You should see all your nodes with READY status.

Run Something

Deploy a Service

sudo swarmcracker service create --name web --image nginx:alpine --replicas 3

See What’s Running

swarmcracker service ls
swarmcracker service ps web

Each task is a Firecracker microVM.

What’s Actually Happening

Manager (swarmd)
    │
    │ gRPC: schedules tasks, maintains state
    │
┌───┴───────────────────┐
│                       │
Worker-1              Worker-2
swarm-br0             swarm-br0
┌───┐┌───┐            ┌───┐┌───┐
│VM1││VM2│  ← VXLAN → │VM3││VM4│
└───┘└───┘            └───┘└───┘

Manager runs SwarmKit control plane
Workers run swarmd-firecracker, which turns SwarmKit tasks into microVMs
swarm-br0 is a Linux bridge for local VM networking
VXLAN connects VMs across different nodes

Common Problems

Kernel: Invalid ELF Magic Number

The kernel file isn’t actually a kernel. Probably HTML from a bad download.

file /usr/share/firecracker/vmlinux
# If it says "HTML document", re-download:
sudo swarmcracker setup install --download-kernel
# Or extract from your host kernel:
./test-automation/scripts/extract-vmlinux.sh /boot/vmlinuz-* /usr/share/firecracker/vmlinux

KVM Not Found

sudo modprobe kvm_intel   # Intel
sudo modprobe kvm_amd     # AMD

Nested KVM Issues

Running inside a VM? Check:

cat /sys/module/kvm_intel/parameters/nested

# If it's 'N':
sudo modprobe -r kvm_intel
sudo modprobe kvm_intel nested=1

Or add options kvm_intel nested=1 to /etc/modprobe.d/kvm-nested.conf.

Workers Can’t Connect

curl http://<manager-ip>:4242   # Check manager reachable
sudo swarmcracker node ls

If the manager advertises 0.0.0.0:4242, that’s wrong. Re-init with --advertise-addr <actual-ip>:4242.

Services Not Starting

sudo swarmcracker node ls             # Check nodes are ready
sudo swarmcracker doctor              # Diagnose common issues
sudo journalctl -u swarmcracker-manager -f   # Manager logs
sudo journalctl -u swarmcracker-worker -f    # Worker logs
file /usr/share/firecracker/vmlinux   # Verify kernel is ELF

Configuration — More options
Networking — VXLAN setup
Security — Jailer hardening
CLI Reference — All commands