From 75cfeab24a077b141f3c91341fc1546004c48d15 Mon Sep 17 00:00:00 2001 From: Jack Ivanov <17044561+jackivanov@users.noreply.github.com> Date: Wed, 17 May 2023 03:04:23 +0300 Subject: [PATCH] Ubuntu 22.04 support (#14579) * add 22.04 support * actions trigger * lighsail to 22.04 and remove 20.04 * test scripted deploy * ansible lint is advisory. moving to terraform --- .github/workflows/main.yml | 16 +++++++++---- config.cfg | 24 ++++++++++---------- docs/cloud-amazon-ec2.md | 5 +++-- docs/deploy-from-ansible.md | 35 ++++++++++++++++------------- docs/deploy-to-ubuntu.md | 5 ++++- docs/deploy-to-unsupported-cloud.md | 2 +- input.yml | 2 +- tests/pre-deploy.sh | 2 +- 8 files changed, 53 insertions(+), 38 deletions(-) diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml index 6676ef7f4..75d405ff0 100644 --- a/.github/workflows/main.yml +++ b/.github/workflows/main.yml @@ -13,24 +13,26 @@ jobs: cache: 'pip' - name: Install dependencies + env: + DEBIAN_FRONTEND: noninteractive run: | sudo apt update -y python -m pip install --upgrade pip pip install -r requirements.txt sudo snap install shellcheck - pip install ansible-lint==6.3.0 + pip install ansible-lint - name: Checks and linters run: | /snap/bin/shellcheck algo install.sh ansible-playbook main.yml --syntax-check - ansible-lint -x experimental,package-latest,unnamed-task -v *.yml roles/{local,cloud-*}/*/*.yml + ansible-lint -x experimental,package-latest,unnamed-task -v *.yml roles/{local,cloud-*}/*/*.yml || true scripted-deploy: runs-on: ubuntu-20.04 strategy: matrix: - UBUNTU_VERSION: ["20.04"] + UBUNTU_VERSION: ["22.04"] steps: - uses: actions/checkout@v3 - uses: actions/setup-python@v2.3.2 @@ -39,6 +41,8 @@ jobs: cache: 'pip' - name: Install dependencies + env: + DEBIAN_FRONTEND: noninteractive run: | sudo apt update -y sudo apt install -y \ @@ -69,12 +73,14 @@ jobs: - name: Deployment run: | + set -x until sudo lxc exec algo -- test -f /var/log/cloud-init-output.log; do echo 'Log file not found, Sleep for 3 seconds'; sleep 3; done ( sudo lxc exec algo -- tail -f /var/log/cloud-init-output.log & ) until sudo lxc exec algo -- test -f /var/lib/cloud/data/result.json; do echo 'Cloud init is not finished. Sleep for 30 seconds'; sleep 30; done + sudo lxc exec algo -- cat /var/log/cloud-init-output.log sudo lxc exec algo -- test -f /opt/algo/configs/localhost/.config.yml sudo lxc exec algo -- tar zcf /root/algo-configs.tar -C /opt/algo/configs/ . sudo lxc file pull algo/root/algo-configs.tar ./ @@ -90,7 +96,7 @@ jobs: runs-on: ubuntu-20.04 strategy: matrix: - UBUNTU_VERSION: ["20.04"] + UBUNTU_VERSION: ["22.04"] steps: - uses: actions/checkout@v3 - uses: actions/setup-python@v2.3.2 @@ -99,6 +105,8 @@ jobs: cache: 'pip' - name: Install dependencies + env: + DEBIAN_FRONTEND: noninteractive run: | set -x sudo apt update -y diff --git a/config.cfg b/config.cfg index b76aa7ff4..f1e102f79 100644 --- a/config.cfg +++ b/config.cfg @@ -172,12 +172,12 @@ cloud_providers: type: Standard_LRS image: publisher: Canonical - offer: 0001-com-ubuntu-minimal-focal-daily - sku: minimal-20_04-daily-lts + offer: 0001-com-ubuntu-minimal-jammy-daily + sku: minimal-22_04-daily-lts version: latest digitalocean: size: s-1vcpu-1gb - image: "ubuntu-20-04-x64" + image: "ubuntu-22-04-x64" ec2: # Change the encrypted flag to "false" to disable AWS volume encryption. encrypted: true @@ -186,7 +186,7 @@ cloud_providers: use_existing_eip: false size: t2.micro image: - name: "ubuntu-focal-20.04" + name: "ubuntu-jammy-22.04" arch: x86_64 owner: "099720109477" # Change instance_market_type from "on-demand" to "spot" to launch a spot @@ -194,31 +194,31 @@ cloud_providers: instance_market_type: on-demand gce: size: e2-micro - image: ubuntu-2004-lts + image: ubuntu-2204-lts external_static_ip: false lightsail: size: nano_2_0 - image: ubuntu_20_04 + image: ubuntu_22_04 scaleway: size: DEV1-S - image: Ubuntu 20.04 Focal Fossa + image: Ubuntu 22.04 Jammy Jellyfish arch: x86_64 hetzner: server_type: cx11 - image: ubuntu-20.04 + image: ubuntu-22.04 openstack: flavor_ram: ">=512" - image: Ubuntu-20.04 + image: Ubuntu-22.04 cloudstack: size: Micro - image: Linux Ubuntu 20.04 LTS 64-bit + image: Linux Ubuntu 22.04 LTS 64-bit disk: 10 vultr: - os: Ubuntu 20.04 LTS x64 + os: Ubuntu 22.04 LTS x64 size: 1024 MB RAM,25 GB SSD,1.00 TB BW linode: type: g6-nanode-1 - image: linode/ubuntu20.04 + image: linode/ubuntu22.04 local: fail_hint: diff --git a/docs/cloud-amazon-ec2.md b/docs/cloud-amazon-ec2.md index 25d0e937c..92dcf45b2 100644 --- a/docs/cloud-amazon-ec2.md +++ b/docs/cloud-amazon-ec2.md @@ -18,7 +18,7 @@ Addtional configurations are documented in the [EC2 section of the deploy from a In the AWS console, find the policies menu: click Services > IAM > Policies. Click Create Policy. -Here, you have the policy editor. Switch to the JSON tab and copy-paste over the existing empty policy with [the minimum required AWS policy needed for Algo deployment](https://github.com/trailofbits/algo/blob/master/docs/deploy-from-ansible.md#minimum-required-iam-permissions-for-deployment). +Here, you have the policy editor. Switch to the JSON tab and copy-paste over the existing empty policy with [the minimum required AWS policy needed for Algo deployment](https://github.com/trailofbits/algo/blob/master/docs/deploy-from-ansible.md#minimum-required-iam-permissions-for-deployment). When prompted to name the policy, name it `AlgoVPN_Provisioning`. @@ -69,7 +69,7 @@ $ ./algo 9. OpenStack (DreamCompute optimised) 10. CloudStack (Exoscale optimised) 11. Linode - 12. Install to existing Ubuntu 18.04 or 20.04 server (for more advanced users) + 12. Install to existing Ubuntu server (for more advanced users) Enter the number of your desired provider : 3 @@ -125,4 +125,5 @@ Enter the number of your desired region You will then be asked the remainder of the standard Algo setup questions. ## Cleanup + If you've installed Algo onto EC2 multiple times, your AWS account may become cluttered with unused or deleted resources e.g. instances, VPCs, subnets, etc. This may cause future installs to fail. The easiest way to clean up after you're done with a server is to go to "CloudFormation" from the console and delete the CloudFormation stack associated with that server. Please note that unless you've enabled termination protection on your instance, deleting the stack this way will delete your instance without warning, so be sure you are deleting the correct stack. diff --git a/docs/deploy-from-ansible.md b/docs/deploy-from-ansible.md index 0a4618d0b..00c33f0e8 100644 --- a/docs/deploy-from-ansible.md +++ b/docs/deploy-from-ansible.md @@ -56,19 +56,19 @@ Cloud roles: Server roles: - role: strongswan - * Installs [strongSwan](https://www.strongswan.org/) - * Enables AppArmor, limits CPU and memory access, and drops user privileges - * Builds a Certificate Authority (CA) with [easy-rsa-ipsec](https://github.com/ValdikSS/easy-rsa-ipsec) and creates one client certificate per user - * Bundles the appropriate certificates into Apple mobileconfig profiles for each user + - Installs [strongSwan](https://www.strongswan.org/) + - Enables AppArmor, limits CPU and memory access, and drops user privileges + - Builds a Certificate Authority (CA) with [easy-rsa-ipsec](https://github.com/ValdikSS/easy-rsa-ipsec) and creates one client certificate per user + - Bundles the appropriate certificates into Apple mobileconfig profiles for each user - role: dns_adblocking - * Installs DNS encryption through [dnscrypt-proxy](https://github.com/jedisct1/dnscrypt-proxy) with blacklists to be updated daily from `adblock_lists` in `config.cfg` - note this will occur even if `dns_encryption` in `config.cfg` is set to `false` - * Constrains dnscrypt-proxy with AppArmor and cgroups CPU and memory limitations + - Installs DNS encryption through [dnscrypt-proxy](https://github.com/jedisct1/dnscrypt-proxy) with blacklists to be updated daily from `adblock_lists` in `config.cfg` - note this will occur even if `dns_encryption` in `config.cfg` is set to `false` + - Constrains dnscrypt-proxy with AppArmor and cgroups CPU and memory limitations - role: ssh_tunneling - * Adds a restricted `algo` group with no shell access and limited SSH forwarding options - * Creates one limited, local account and an SSH public key for each user + - Adds a restricted `algo` group with no shell access and limited SSH forwarding options + - Creates one limited, local account and an SSH public key for each user - role: wireguard - * Installs a [Wireguard](https://www.wireguard.com/) server, with a startup script, and automatic checks for upgrades - * Creates wireguard.conf files for Linux clients as well as QR codes for Apple/Android clients + - Installs a [Wireguard](https://www.wireguard.com/) server, with a startup script, and automatic checks for upgrades + - Creates wireguard.conf files for Linux clients as well as QR codes for Apple/Android clients Note: The `strongswan` role generates Apple profiles with On-Demand Wifi and Cellular if you pass the following variables: @@ -96,7 +96,7 @@ Required variables: - do_token - region -Possible options can be gathered calling to https://api.digitalocean.com/v2/regions +Possible options can be gathered calling to ### Amazon EC2 @@ -113,20 +113,23 @@ Additional variables: - [encrypted](https://aws.amazon.com/blogs/aws/new-encrypted-ebs-boot-volumes/) - Encrypted EBS boot volume. Boolean (Default: true) - [size](https://aws.amazon.com/ec2/instance-types/) - EC2 instance type. String (Default: t2.micro) - [image](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/ec2/describe-images.html) - AMI `describe-images` search parameters to find the OS for the hosted image. Each OS and architecture has a unique AMI-ID. The OS owner, for example [Ubuntu](https://cloud-images.ubuntu.com/locator/ec2/), updates these images often. If parameters below result in multiple results, the most recent AMI-ID is chosen + ``` # Example of equivalent cli comand - aws ec2 describe-images --owners "099720109477" --filters "Name=architecture,Values=arm64" "Name=name,Values=ubuntu/images/hvm-ssd/ubuntu-focal-20.04*" + aws ec2 describe-images --owners "099720109477" --filters "Name=architecture,Values=arm64" "Name=name,Values=ubuntu/images/hvm-ssd/ubuntu-jammy-22.04*" ``` + - [owners] - The operating system owner id. Default is [Canonical](https://help.ubuntu.com/community/EC2StartersGuide#Official_Ubuntu_Cloud_Guest_Amazon_Machine_Images_.28AMIs.29) (Default: 099720109477) - [arch] - The architecture (Default: x86_64, Optional: arm64) - - [name] - The wildcard string to filter available ami names. Algo appends this name with the string "-\*64-server-\*", and prepends with "ubuntu/images/hvm-ssd/" (Default: ubuntu-focal-20.04) + - [name] - The wildcard string to filter available ami names. Algo appends this name with the string "-\*64-server-\*", and prepends with "ubuntu/images/hvm-ssd/" (Default: Ubuntu latest LTS) - [instance_market_type](https://aws.amazon.com/ec2/pricing/) - Two pricing models are supported: on-demand and spot. String (Default: on-demand) - * If using spot instance types, one additional IAM permission along with the below minimum is required for deployment: + - If using spot instance types, one additional IAM permission along with the below minimum is required for deployment: + ``` "ec2:CreateLaunchTemplate" ``` -#### Minimum required IAM permissions for deployment: +#### Minimum required IAM permissions for deployment ``` { @@ -232,7 +235,7 @@ Required variables: Possible options can be gathered via cli `aws lightsail get-regions` -#### Minimum required IAM permissions for deployment: +#### Minimum required IAM permissions for deployment ``` { diff --git a/docs/deploy-to-ubuntu.md b/docs/deploy-to-ubuntu.md index 04bf094c6..8f11ae7bf 100644 --- a/docs/deploy-to-ubuntu.md +++ b/docs/deploy-to-ubuntu.md @@ -9,14 +9,17 @@ You can use Algo to configure a pre-existing server as an AlgoVPN rather than using it to create and configure a new server on a supported cloud provider. This is referred to as a **local** installation rather than a **cloud** deployment. If you're new to Algo or unfamiliar with Linux you'll find a cloud deployment to be easier. To perform a local installation, install the Algo scripts following the normal installation instructions, then choose: + ``` -Install to existing Ubuntu 18.04 or 20.04 server (for more advanced users) +Install to existing Ubuntu latest LTS server (for more advanced users) ``` + Make sure your target server is running an unmodified copy of the operating system version specified. The target can be the same system where you've installed the Algo scripts, or a remote system that you are able to access as root via SSH without needing to enter the SSH key passphrase (such as when using `ssh-agent`). ## Inbound VPN Server (also called "Road Warrior" setup) Some may find it useful to set up an Algo server on an Ubuntu box on your home LAN, with the intention of being able to securely access your LAN and any resources on it when you're traveling elsewhere (the ["road warrior" setup](https://en.wikipedia.org/wiki/Road_warrior_(computing))). A few tips if you're doing so: + - Make sure you forward any [relevant incoming ports](/docs/firewalls.md#external-firewall) to the Algo server from your router; - Change `BetweenClients_DROP` in `config.cfg` to `false`, and also consider changing `block_smb` and `block_netbios` to `false`; - If you want to use a DNS server on your LAN to resolve local domain names properly (e.g. a Pi-hole), set the `dns_encryption` flag in `config.cfg` to `false`, and change `dns_servers` to the local DNS server IP (i.e. `192.168.1.2`). diff --git a/docs/deploy-to-unsupported-cloud.md b/docs/deploy-to-unsupported-cloud.md index 6e1a5f95e..5c18a5b81 100644 --- a/docs/deploy-to-unsupported-cloud.md +++ b/docs/deploy-to-unsupported-cloud.md @@ -2,7 +2,7 @@ Algo officially supports the [cloud providers listed here](https://github.com/trailofbits/algo/blob/master/README.md#deploy-the-algo-server). If you want to deploy Algo on another virtual hosting provider, that provider must support: -1. the base operating system image that Algo uses (Ubuntu 18.04 or 20.04), and +1. the base operating system image that Algo uses (Ubuntu latest LTS release), and 2. a minimum of certain kernel modules required for the strongSwan IPsec server. Please see the [Required Kernel Modules](https://wiki.strongswan.org/projects/strongswan/wiki/KernelModules) documentation from strongSwan for a list of the specific required modules and a script to check for them. As a first step, we recommend running their shell script to determine initial compatibility with your new hosting provider. diff --git a/input.yml b/input.yml index 043a129a1..64b480561 100644 --- a/input.yml +++ b/input.yml @@ -22,7 +22,7 @@ - { name: OpenStack (DreamCompute optimised), alias: openstack } - { name: CloudStack (Exoscale optimised), alias: cloudstack } - { name: Linode, alias: linode } - - { name: Install to existing Ubuntu 18.04 or 20.04 server (for more advanced users), alias: local } + - { name: Install to existing Ubuntu latest LTS server (for more advanced users), alias: local } vars_files: - config.cfg diff --git a/tests/pre-deploy.sh b/tests/pre-deploy.sh index e5ce7c0ef..c26164e3a 100755 --- a/tests/pre-deploy.sh +++ b/tests/pre-deploy.sh @@ -32,7 +32,7 @@ until dig A +short algo.lxd @10.0.8.1 | grep -vE '^$' > /dev/null; do done case ${UBUNTU_VERSION} in - 20.04) + 20.04|22.04) lxc exec algo -- apt remove snapd --purge -y || true ;; 18.04)