Skip to content

Latest commit

 

History

History
390 lines (288 loc) · 12.4 KB

getting-started.md

File metadata and controls

390 lines (288 loc) · 12.4 KB
Raw

Getting Started with Firecracker

All resources are used for demonstration purposes and are not intended for production.

Prerequisites

You can check if your system meets the requirements by running firecracker/tools/devtool checkenv.

An opinionated way to run Firecracker is to launch an EC2 c5.metal instance with Ubuntu 24.04.

Firecracker requires the KVM Linux kernel module to perform its virtualization and emulation tasks.

We exclusively use .metal instance types, because EC2 only supports KVM on .metal instance types.

Architecture & OS

Firecracker supports x86_64 and aarch64 Linux, see specific supported kernels.

KVM

Firecracker requires read/write access to /dev/kvm exposed by the KVM module.

The presence of the KVM module can be checked with:

lsmod | grep kvm

An example output where it is enabled:

kvm_intel             348160  0
kvm                   970752  1 kvm_intel
irqbypass              16384  1 kvm

Some Linux distributions use the kvm group to manage access to /dev/kvm, while others rely on access control lists. If you have the ACL package for your distro installed, you can grant Read+Write access with:

sudo setfacl -m u:${USER}:rw /dev/kvm

If access is managed via the kvm group, check that the KVM group exists:

getent group kvm

and check that /dev/kvm is associated with the kvm group:

ls -l /dev/kvm

You can see if your current user is already in the kvm group by running:

groups

Otherwise, add your current user to the group by running:

[ $(stat -c "%G" /dev/kvm) = kvm ] && sudo usermod -aG kvm ${USER} \
&& echo "Access granted."

If none of the above works, you will need to either install the file system ACL package for your distro and use the setfacl command as above, or run Firecracker as root (via sudo).

You can check if you have access to /dev/kvm with:

[ -r /dev/kvm ] && [ -w /dev/kvm ] && echo "OK" || echo "FAIL"

Running Firecracker

In production, Firecracker is designed to be run securely inside an execution jail, set up by the jailer binary. This is how our integration test suite does it.

For simplicity, this guide will not use the jailer.

Getting a rootfs and Guest Kernel Image

To successfully start a microVM, you will need an uncompressed Linux kernel binary, and an ext4 file system image (to use as rootfs). This guide uses a 5.10 kernel image with a Ubuntu 24.04 rootfs from our CI:

ARCH="$(uname -m)"

latest=$(wget "http://spec.ccfc.min.s3.amazonaws.com/?prefix=firecracker-ci/v1.11/$ARCH/vmlinux-5.10&list-type=2" -O - 2>/dev/null | grep -oP "(?<=<Key>)(firecracker-ci/v1.11/$ARCH/vmlinux-5\.10\.[0-9]{1,3})(?=</Key>)")

# Download a linux kernel binary
wget "https://s3.amazonaws.com/spec.ccfc.min/${latest}"

# Download a rootfs
wget -O ubuntu-24.04.squashfs.upstream "https://s3.amazonaws.com/spec.ccfc.min/firecracker-ci/v1.11/${ARCH}/ubuntu-24.04.squashfs"

# Create an ssh key for the rootfs
unsquashfs ubuntu-24.04.squashfs.upstream
ssh-keygen -f id_rsa -N ""
cp -v id_rsa.pub squashfs-root/root/.ssh/authorized_keys
mv -v id_rsa ./ubuntu-24.04.id_rsa
# create ext4 filesystem image
sudo chown -R root:root squashfs-root
truncate -s 400M ubuntu-24.04.ext4
sudo mkfs.ext4 -d squashfs-root -F ubuntu-24.04.ext4

Getting a Firecracker Binary

There are two options for getting a firecracker binary:

  • Downloading an official firecracker release from our release page, or
  • Building firecracker from source.

To download the latest firecracker release, run:

ARCH="$(uname -m)"
release_url="https://github.com/firecracker-microvm/firecracker/releases"
latest=$(basename $(curl -fsSLI -o /dev/null -w  %{url_effective} ${release_url}/latest))
curl -L ${release_url}/download/${latest}/firecracker-${latest}-${ARCH}.tgz \
| tar -xz

# Rename the binary to "firecracker"
mv release-${latest}-$(uname -m)/firecracker-${latest}-${ARCH} firecracker

To instead build firecracker from source, you will need to have docker installed:

ARCH="$(uname -m)"

# Clone the firecracker repository
git clone https://github.com/firecracker-microvm/firecracker firecracker_src

# Start docker
sudo systemctl start docker

# Build firecracker
#
# It is possible to build for gnu, by passing the arguments '-l gnu'.
#
# This will produce the firecracker and jailer binaries under
# `./firecracker/build/cargo_target/${toolchain}/debug`.
#
sudo ./firecracker_src/tools/devtool build

# Rename the binary to "firecracker"
sudo cp ./firecracker_src/build/cargo_target/${ARCH}-unknown-linux-musl/debug/firecracker firecracker

Starting Firecracker

Running firecracker will require two terminals, the first one running the firecracker binary, and a second one for communicating with the firecracker process via HTTP requests:

API_SOCKET="/tmp/firecracker.socket"

# Remove API unix socket
sudo rm -f $API_SOCKET

# Run firecracker
sudo ./firecracker --api-sock "${API_SOCKET}"

In a new terminal (do not close the 1st one):

TAP_DEV="tap0"
TAP_IP="172.16.0.1"
MASK_SHORT="/30"

# Setup network interface
sudo ip link del "$TAP_DEV" 2> /dev/null || true
sudo ip tuntap add dev "$TAP_DEV" mode tap
sudo ip addr add "${TAP_IP}${MASK_SHORT}" dev "$TAP_DEV"
sudo ip link set dev "$TAP_DEV" up

# Enable ip forwarding
sudo sh -c "echo 1 > /proc/sys/net/ipv4/ip_forward"
sudo iptables -P FORWARD ACCEPT

# This tries to determine the name of the host network interface to forward
# VM's outbound network traffic through. If outbound traffic doesn't work,
# double check this returns the correct interface!
HOST_IFACE=$(ip -j route list default |jq -r '.[0].dev')

# Set up microVM internet access
sudo iptables -t nat -D POSTROUTING -o "$HOST_IFACE" -j MASQUERADE || true
sudo iptables -t nat -A POSTROUTING -o "$HOST_IFACE" -j MASQUERADE

API_SOCKET="/tmp/firecracker.socket"
LOGFILE="./firecracker.log"

# Create log file
touch $LOGFILE

# Set log file
sudo curl -X PUT --unix-socket "${API_SOCKET}" \
    --data "{
        \"log_path\": \"${LOGFILE}\",
        \"level\": \"Debug\",
        \"show_level\": true,
        \"show_log_origin\": true
    }" \
    "http://localhost/logger"

KERNEL="./$(ls vmlinux* | tail -1)"
KERNEL_BOOT_ARGS="console=ttyS0 reboot=k panic=1 pci=off"

ARCH=$(uname -m)

if [ ${ARCH} = "aarch64" ]; then
    KERNEL_BOOT_ARGS="keep_bootcon ${KERNEL_BOOT_ARGS}"
fi

# Set boot source
sudo curl -X PUT --unix-socket "${API_SOCKET}" \
    --data "{
        \"kernel_image_path\": \"${KERNEL}\",
        \"boot_args\": \"${KERNEL_BOOT_ARGS}\"
    }" \
    "http://localhost/boot-source"

ROOTFS="./ubuntu-24.04.ext4"

# Set rootfs
sudo curl -X PUT --unix-socket "${API_SOCKET}" \
    --data "{
        \"drive_id\": \"rootfs\",
        \"path_on_host\": \"${ROOTFS}\",
        \"is_root_device\": true,
        \"is_read_only\": false
    }" \
    "http://localhost/drives/rootfs"

# The IP address of a guest is derived from its MAC address with
# `fcnet-setup.sh`, this has been pre-configured in the guest rootfs. It is
# important that `TAP_IP` and `FC_MAC` match this.
FC_MAC="06:00:AC:10:00:02"

# Set network interface
sudo curl -X PUT --unix-socket "${API_SOCKET}" \
    --data "{
        \"iface_id\": \"net1\",
        \"guest_mac\": \"$FC_MAC\",
        \"host_dev_name\": \"$TAP_DEV\"
    }" \
    "http://localhost/network-interfaces/net1"

# API requests are handled asynchronously, it is important the configuration is
# set, before `InstanceStart`.
sleep 0.015s

# Start microVM
sudo curl -X PUT --unix-socket "${API_SOCKET}" \
    --data "{
        \"action_type\": \"InstanceStart\"
    }" \
    "http://localhost/actions"

# API requests are handled asynchronously, it is important the microVM has been
# started before we attempt to SSH into it.
sleep 2s

# Setup internet access in the guest
ssh -i ./ubuntu-24.04.id_rsa [email protected]  "ip route add default via 172.16.0.1 dev eth0"

# Setup DNS resolution in the guest
ssh -i ./ubuntu-24.04.id_rsa [email protected]  "echo 'nameserver 8.8.8.8' > /etc/resolv.conf"

# SSH into the microVM
ssh -i ./ubuntu-24.04.id_rsa [email protected]

# Use `root` for both the login and password.
# Run `reboot` to exit.

Issuing a reboot command inside the guest will gracefully shutdown Firecracker. This is due to the fact that Firecracker doesn't implement guest power management.

Configuring the microVM without sending API requests

You can boot a guest without using the API socket by passing the parameter --config-file to the Firecracker process. E.g.:

sudo ./firecracker --api-sock /tmp/firecracker.socket --config-file <path_to_the_configuration_file>

path_to_the_configuration_file is the path to a JSON file with the configuration for all of the microVM's resources. The JSON must contain the configuration for the guest kernel and rootfs, all of the other resources are optional. This configuration method will also start the microVM, as such you need to specify all desired pre-boot configurable resources in the JSON. The names of the resources can be seen in [firecracker.yaml] (../src/firecracker/swagger/firecracker.yaml) and the names of their fields are the same that are used in the API requests.

An example of configuration file is provided: tests/framework/vm_config.json.

Once the guest is booted, refer network-setup to bring up the network in the guest machine.

After the microVM is started you can still use the socket to send API requests for post-boot operations.

Building Firecracker

SSH can be used to work with libraries from private git repos by passing the --ssh-keys flag to specify the paths to your public and private SSH keys on the host. Both are required for git authentication when fetching the repositories.

tools/devtool build --ssh-keys ~/.ssh/id_rsa.pub ~/.ssh/id_rsa

Only a single set of credentials is supported. devtool cannot fetch multiple private repos which rely on different credentials.

tools/devtool build builds in debug to build release binaries pass --release e.g. tools/devtool build --release

Documentation on devtool can be seen with tools/devtool --help.

Running the Integration Test Suite

Integration tests can be run with tools/devtool test.

The test suite is designed to ensure our SLA parameters as measured on EC2 .metal instances, as such performance tests may fail when not run on these machines. Specifically, don't be alarmed if you see tests/integration_tests/performance/test_process_startup_time.py failing when not run on an EC2 .metal instance. You can skip performance tests with:

./tools/devtool test -- --ignore integration_tests/performance

If you run the integration tests on an EC2 .metal instance, and encounter failures such as the following

FAILED integration_tests/style/test_markdown.py::test_markdown_style - requests.exceptions.ReadTimeout: HTTPConnectionPool(host='169.254.169.254', port=80): Read timed out. (read timeout=2)

try running aws ec2 modify-instance-metadata-options --instance-id i-<your instance id> --http-put-response-hop-limit 2. The integration tests framework uses IMDSv2 to determine information such as instance type. The additional hop is needed because the IMDS requests will pass through docker.

Errors while using curl to access the API

Points to check to confirm the API socket is running and accessible:

  • Check that the user running the Firecracker process and the user using curl have equivalent privileges. For example, if you run Firecracker with sudo that you run curl with sudo as well.
  • SELinux can regulate access to sockets on RHEL based distributions. How user's permissions are configured is environmentally specific, but for the purposes of troubleshooting you can check if it is enabled in /etc/selinux/config.
  • With the Firecracker process running using --api-sock /tmp/firecracker.socket, confirm that the socket is open:
    • ss -a | grep '/tmp/firecracker.socket'
    • If you have socat available, try socat - UNIX-CONNECT:/tmp/firecracker.socket This will throw an explicit error if the socket is inaccessible, or it will pause and wait for input to continue.