torjus/nixos-servers
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects 1 Releases Wiki Activity

pipeline-testing-improvements #9

Merged
torjus merged 7 commits from pipeline-testing-improvements into master 2026-02-01 16:45:04 +00:00
Conversation 0 Commits 7 Files Changed 12 +354 -54
12 changed files with 354 additions and 54 deletions
Download Patch File Download Diff File Expand all files Collapse all files

84
TODO.md
View File

@@ -54,6 +54,7 @@ Automate the entire process of creating, configuring, and deploying new NixOS ho
**Status:** ✅ Fully implemented and tested
**Completed:** 2025-02-01
**Enhanced:** 2025-02-01 (added --force flag)
**Goal:** Automate creation of host configuration files
@@ -64,6 +65,7 @@ Automate the entire process of creating, configuring, and deploying new NixOS ho
- Comprehensive validation (hostname format/uniqueness, IP subnet/uniqueness)
- Jinja2 templates for NixOS configurations
- Automatic updates to flake.nix and terraform/vms.tf
- `--force` flag for regenerating existing configurations (useful for testing)
**Tasks:**
- [x] Create Python CLI with typer framework
@@ -109,6 +111,7 @@ create-host \
**Status:** ✅ Fully implemented and tested
**Completed:** 2025-02-01
**Enhanced:** 2025-02-01 (added branch support for testing)
**Goal:** Get freshly deployed VM to apply its specific host configuration
@@ -118,7 +121,8 @@ create-host \
- Systemd service `nixos-bootstrap.service` runs on first boot
- Depends on `cloud-config.service` to ensure hostname is set
- Reads hostname from `hostnamectl` (set by cloud-init via Terraform)
- Runs `nixos-rebuild boot --flake git+https://git.t-juice.club/torjus/nixos-servers.git#${hostname}`
- Supports custom git branch via `NIXOS_FLAKE_BRANCH` environment variable
- Runs `nixos-rebuild boot --flake git+https://git.t-juice.club/torjus/nixos-servers.git?ref=$BRANCH#${hostname}`
- Reboots into new configuration on success
- Fails gracefully without reboot on errors (network issues, missing config)
- Service self-destructs after successful bootstrap (not in new config)
@@ -240,10 +244,80 @@ Since most hosts use static IPs defined in their NixOS configurations, we can ex
### Phase 7: Testing & Documentation
**Tasks:**
- [ ] Test full pipeline end-to-end
- [ ] Create test host and verify all steps
- [ ] Document the new workflow in CLAUDE.md
**Status:** 🚧 In Progress (testing improvements completed)
**Testing Improvements Implemented (2025-02-01):**
The pipeline now supports efficient testing without polluting master branch:
**1. --force Flag for create-host**
- Re-run `create-host` to regenerate existing configurations
- Updates existing entries in flake.nix and terraform/vms.tf (no duplicates)
- Skip uniqueness validation checks
- Useful for iterating on configuration templates during testing
**2. Branch Support for Bootstrap**
- Bootstrap service reads `NIXOS_FLAKE_BRANCH` environment variable
- Defaults to `master` if not set
- Allows testing pipeline changes on feature branches
- Cloud-init passes branch via `/etc/environment`
**3. Cloud-init Disk for Branch Configuration**
- Terraform generates custom cloud-init snippets for test VMs
- Set `flake_branch` field in VM definition to use non-master branch
- Production VMs omit this field and use master (default)
- Files automatically uploaded to Proxmox via SSH
**Testing Workflow:**
```bash
# 1. Create test branch
git checkout -b test-pipeline
# 2. Generate or update host config
create-host --hostname testvm01 --ip 10.69.13.100/24
# 3. Edit terraform/vms.tf to add test VM with branch
# vms = {
# "testvm01" = {
# ip = "10.69.13.100/24"
# flake_branch = "test-pipeline" # Bootstrap from this branch
# }
# }
# 4. Commit and push test branch
git add -A && git commit -m "test: add testvm01"
git push origin test-pipeline
# 5. Deploy VM
cd terraform && tofu apply
# 6. Watch bootstrap (VM fetches from test-pipeline branch)
ssh root@10.69.13.100
journalctl -fu nixos-bootstrap.service
# 7. Iterate: modify templates and regenerate with --force
cd .. && create-host --hostname testvm01 --ip 10.69.13.100/24 --force
git commit -am "test: update config" && git push
# Redeploy to test fresh bootstrap
cd terraform
tofu destroy -target=proxmox_vm_qemu.vm[\"testvm01\"] && tofu apply
# 8. Clean up when done: squash commits, merge to master, remove test VM
```
**Files:**
- `scripts/create-host/create_host.py` - Added --force parameter
- `scripts/create-host/manipulators.py` - Update vs insert logic
- `hosts/template2/bootstrap.nix` - Branch support via environment variable
- `terraform/vms.tf` - flake_branch field support
- `terraform/cloud-init.tf` - Custom cloud-init disk generation
- `terraform/variables.tf` - proxmox_host variable for SSH uploads
**Remaining Tasks:**
- [ ] Test full pipeline end-to-end on feature branch
- [ ] Update CLAUDE.md with testing workflow
- [ ] Add troubleshooting section
- [ ] Create examples for common scenarios (DHCP host, static IP host, etc.)

16
flake.nix
View File

@@ -334,6 +334,22 @@
sops-nix.nixosModules.sops
];
};
testvm01 = nixpkgs.lib.nixosSystem {
inherit system;
specialArgs = {
inherit inputs self sops-nix;
};
modules = [
(
{ config, pkgs, ... }:
{
nixpkgs.overlays = commonOverlays;
}
)
./hosts/testvm01
sops-nix.nixosModules.sops
];
};
};
packages = forAllSystems (
{ pkgs }:

9
hosts/template2/bootstrap.nix
View File

@@ -24,8 +24,12 @@ let
echo "Network connectivity confirmed"
echo "Fetching and building NixOS configuration from flake..."
# Read git branch from environment, default to master
BRANCH="''${NIXOS_FLAKE_BRANCH:-master}"
echo "Using git branch: $BRANCH"
# Build and activate the host-specific configuration
FLAKE_URL="git+https://git.t-juice.club/torjus/nixos-servers.git#''${HOSTNAME}"
FLAKE_URL="git+https://git.t-juice.club/torjus/nixos-servers.git?ref=$BRANCH#''${HOSTNAME}"
if nixos-rebuild boot --flake "$FLAKE_URL"; then
echo "Successfully built configuration for $HOSTNAME"
@@ -58,6 +62,9 @@ in
RemainAfterExit = true;
ExecStart = "${bootstrap-script}/bin/nixos-bootstrap";
# Read environment variables from /etc/environment (set by cloud-init)
EnvironmentFile = "-/etc/environment";
# Logging to journald
StandardOutput = "journal+console";
StandardError = "journal+console";

61
hosts/testvm01/configuration.nix Normal file
View File

@@ -0,0 +1,61 @@
{
config,
lib,
pkgs,
...
}:
{
imports = [
../template2/hardware-configuration.nix
../../system
../../common/vm
];
nixpkgs.config.allowUnfree = true;
boot.loader.grub.enable = true;
boot.loader.grub.device = "/dev/vda";
networking.hostName = "testvm01";
networking.domain = "home.2rjus.net";
networking.useNetworkd = true;
networking.useDHCP = false;
services.resolved.enable = false;
networking.nameservers = [
"10.69.13.5"
"10.69.13.6"
];
systemd.network.enable = true;
systemd.network.networks."ens18" = {
matchConfig.Name = "ens18";
address = [
"10.69.13.101/24"
];
routes = [
{ Gateway = "10.69.13.1"; }
];
linkConfig.RequiredForOnline = "routable";
};
time.timeZone = "Europe/Oslo";
nix.settings.experimental-features = [
"nix-command"
"flakes"
];
nix.settings.tarball-ttl = 0;
environment.systemPackages = with pkgs; [
vim
wget
git
];
# Open ports in the firewall.
# networking.firewall.allowedTCPPorts = [ ... ];
# networking.firewall.allowedUDPPorts = [ ... ];
# Or disable the firewall altogether.
networking.firewall.enable = false;
system.stateVersion = "25.11"; # Did you read the comment?
}

5
hosts/testvm01/default.nix Normal file
View File

@@ -0,0 +1,5 @@
{ ... }: {
imports = [
./configuration.nix
];
}

21
scripts/create-host/README.md
View File

@@ -50,6 +50,23 @@ python -m scripts.create_host.create_host create \
--dry-run
```
### Force Mode (Regenerate Existing Configuration)
Overwrite an existing host configuration (useful for testing):
```bash
python -m scripts.create_host.create_host create \
--hostname test01 \
--ip 10.69.13.50/24 \
--force
```
This mode:
- Skips hostname and IP uniqueness validation
- Overwrites files in `hosts/<hostname>/`
- Updates existing entries in `flake.nix` and `terraform/vms.tf` (doesn't duplicate)
- Useful for iterating on configuration templates during testing
### Options
- `--hostname` (required): Hostname for the new host
@@ -73,6 +90,10 @@ python -m scripts.create_host.create_host create \
- `--dry-run` (flag): Preview changes without creating files
- `--force` (flag): Overwrite existing host configuration
- Skips uniqueness validation
- Updates existing entries instead of creating duplicates
## What It Does
The tool performs the following actions:

21
scripts/create-host/create_host.py
View File

@@ -45,6 +45,7 @@ def main(
memory: int = typer.Option(2048, "--memory", help="Memory in MB"),
disk: str = typer.Option("20G", "--disk", help="Disk size (e.g., 20G, 50G, 100G)"),
dry_run: bool = typer.Option(False, "--dry-run", help="Preview changes without creating files"),
force: bool = typer.Option(False, "--force", help="Overwrite existing host configuration"),
) -> None:
"""
Create a new NixOS host configuration.
@@ -75,11 +76,20 @@ def main(
config.validate()
validate_hostname_format(hostname)
# Skip uniqueness checks in force mode
if not force:
validate_hostname_unique(hostname, repo_root)
if ip:
validate_ip_unique(ip, repo_root)
else:
# Check if we're actually overwriting something
host_dir = repo_root / "hosts" / hostname
if host_dir.exists():
console.print(f"[yellow]⚠[/yellow] Updating existing host configuration for {hostname}")
if ip:
validate_ip_subnet(ip)
validate_ip_unique(ip, repo_root)
console.print("[green]✓[/green] All validations passed\n")
@@ -96,13 +106,14 @@ def main(
console.print("\n[bold blue]Generating host configuration...[/bold blue]")
generate_host_files(config, repo_root)
console.print(f"[green]✓[/green] Created hosts/{hostname}/default.nix")
console.print(f"[green]✓[/green] Created hosts/{hostname}/configuration.nix")
action = "Updated" if force else "Created"
console.print(f"[green]✓[/green] {action} hosts/{hostname}/default.nix")
console.print(f"[green]✓[/green] {action} hosts/{hostname}/configuration.nix")
update_flake_nix(config, repo_root)
update_flake_nix(config, repo_root, force=force)
console.print("[green]✓[/green] Updated flake.nix")
update_terraform_vms(config, repo_root)
update_terraform_vms(config, repo_root, force=force)
console.print("[green]✓[/green] Updated terraform/vms.tf")
# Success message

54
scripts/create-host/manipulators.py
View File

@@ -6,21 +6,18 @@ from pathlib import Path
from models import HostConfig
def update_flake_nix(config: HostConfig, repo_root: Path) -> None:
def update_flake_nix(config: HostConfig, repo_root: Path, force: bool = False) -> None:
"""
Add new host entry to flake.nix nixosConfigurations.
Add or update host entry in flake.nix nixosConfigurations.
Args:
config: Host configuration
repo_root: Path to repository root
force: If True, replace existing entry; if False, insert new entry
"""
flake_path = repo_root / "flake.nix"
content = flake_path.read_text()
# Find the closing of nixosConfigurations block
# Pattern: " };\n packages ="
pattern = r"( \};)\n( packages =)"
# Create new entry
new_entry = f""" {config.hostname} = nixpkgs.lib.nixosSystem {{
inherit system;
@@ -40,35 +37,47 @@ def update_flake_nix(config: HostConfig, repo_root: Path) -> None:
}};
"""
# Insert new entry before closing brace
replacement = rf"\g<1>\n{new_entry}\g<2>"
# Check if hostname already exists
hostname_pattern = rf"^ {re.escape(config.hostname)} = nixpkgs\.lib\.nixosSystem"
existing_match = re.search(hostname_pattern, content, re.MULTILINE)
if existing_match and force:
# Replace existing entry
# Match the entire block from "hostname = " to "};"
replace_pattern = rf"^ {re.escape(config.hostname)} = nixpkgs\.lib\.nixosSystem \{{.*?^ \}};\n"
new_content, count = re.subn(replace_pattern, new_entry, content, flags=re.MULTILINE | re.DOTALL)
if count == 0:
raise ValueError(f"Could not find existing entry for {config.hostname} in flake.nix")
else:
# Insert new entry before closing brace of nixosConfigurations
# Pattern: " };\n packages = forAllSystems"
pattern = r"( \};)\n( packages = forAllSystems)"
replacement = rf"{new_entry}\g<1>\n\g<2>"
new_content, count = re.subn(pattern, replacement, content)
if count == 0:
raise ValueError(
"Could not find insertion point in flake.nix. "
"Looking for pattern: ' };\\n devShells ='"
"Looking for pattern: ' };\\n packages = forAllSystems'"
)
flake_path.write_text(new_content)
def update_terraform_vms(config: HostConfig, repo_root: Path) -> None:
def update_terraform_vms(config: HostConfig, repo_root: Path, force: bool = False) -> None:
"""
Add new VM entry to terraform/vms.tf locals.vms map.
Add or update VM entry in terraform/vms.tf locals.vms map.
Args:
config: Host configuration
repo_root: Path to repository root
force: If True, replace existing entry; if False, insert new entry
"""
terraform_path = repo_root / "terraform" / "vms.tf"
content = terraform_path.read_text()
# Find the closing of locals.vms block
# Pattern: " }\n\n # Compute VM configurations"
pattern = r"( \})\n\n( # Compute VM configurations)"
# Create new entry based on whether we have static IP or DHCP
if config.is_static_ip:
new_entry = f''' "{config.hostname}" = {{
@@ -86,7 +95,22 @@ def update_terraform_vms(config: HostConfig, repo_root: Path) -> None:
}}
'''
# Check if hostname already exists
hostname_pattern = rf'^\s+"{re.escape(config.hostname)}" = \{{'
existing_match = re.search(hostname_pattern, content, re.MULTILINE)
if existing_match and force:
# Replace existing entry
# Match the entire block from "hostname" = { to }
replace_pattern = rf'^\s+"{re.escape(config.hostname)}" = \{{.*?^\s+\}}\n'
new_content, count = re.subn(replace_pattern, new_entry, content, flags=re.MULTILINE | re.DOTALL)
if count == 0:
raise ValueError(f"Could not find existing entry for {config.hostname} in terraform/vms.tf")
else:
# Insert new entry before closing brace
# Pattern: " }\n\n # Compute VM configurations"
pattern = r"( \})\n\n( # Compute VM configurations)"
replacement = rf"{new_entry}\g<1>\n\n\g<2>"
new_content, count = re.subn(pattern, replacement, content)

5
scripts/create-host/templates/configuration.nix.j2
View File

@@ -7,16 +7,15 @@
{
imports = [
../template/hardware-configuration.nix
../template2/hardware-configuration.nix
../../system
../../common/vm
];
nixpkgs.config.allowUnfree = true;
# Use the systemd-boot EFI boot loader.
boot.loader.grub.enable = true;
boot.loader.grub.device = "/dev/sda";
boot.loader.grub.device = "/dev/vda";
networking.hostName = "{{ hostname }}";
networking.domain = "{{ domain }}";

17
terraform/README.md
View File

@@ -87,6 +87,21 @@ vms = {
}
```
### Example: Test VM with Custom Git Branch
For testing pipeline changes without polluting master:
```hcl
vms = {
"test-vm" = {
ip = "10.69.13.100/24"
flake_branch = "test-pipeline" # Bootstrap from this branch
}
}
```
This VM will bootstrap from the `test-pipeline` branch instead of `master`. Production VMs should omit the `flake_branch` field.
## Configuration Options
Each VM in the `vms` map supports the following fields (all optional):
@@ -98,6 +113,7 @@ Each VM in the `vms` map supports the following fields (all optional):
| `cpu_cores` | Number of CPU cores | `2` |
| `memory` | Memory in MB | `2048` |
| `disk_size` | Disk size (e.g., "20G", "100G") | `"20G"` |
| `flake_branch` | Git branch for bootstrap (for testing, omit for production) | `master` |
| `target_node` | Proxmox node to deploy to | `"pve1"` |
| `template_name` | Template VM to clone from | `"nixos-25.11.20260128.fa83fd8"` |
| `storage` | Storage backend | `"local-zfs"` |
@@ -182,6 +198,7 @@ deployment_summary = {
- `main.tf` - Provider configuration
- `variables.tf` - Variable definitions and defaults
- `vms.tf` - VM definitions and deployment logic
- `cloud-init.tf` - Cloud-init disk management (SSH keys, networking, branch config)
- `outputs.tf` - Output definitions for deployed VMs
- `terraform.tfvars.example` - Example credentials file
- `terraform.tfvars` - Your actual credentials (gitignored)

58
terraform/cloud-init.tf Normal file
View File

@@ -0,0 +1,58 @@
# Cloud-init configuration for all VMs
#
# This file manages cloud-init disks for all VMs using the proxmox_cloud_init_disk resource.
# VMs with flake_branch set will include NIXOS_FLAKE_BRANCH environment variable.
resource "proxmox_cloud_init_disk" "ci" {
for_each = local.vm_configs
name = each.key
pve_node = each.value.target_node
storage = "local" # Cloud-init disks must be on storage that supports ISO/snippets
# User data includes SSH keys and optionally NIXOS_FLAKE_BRANCH
user_data = <<-EOT
#cloud-config
ssh_authorized_keys:
- ${each.value.ssh_public_key}
${each.value.flake_branch != null ? <<-BRANCH
write_files:
- path: /etc/environment
content: |
NIXOS_FLAKE_BRANCH=${each.value.flake_branch}
append: true
BRANCH
: ""}
EOT
# Network configuration - static IP or DHCP
network_config = each.value.ip != null ? yamlencode({
version = 1
config = [{
type = "physical"
name = "ens18"
subnets = [{
type = "static"
address = each.value.ip
gateway = each.value.gateway
dns_nameservers = split(" ", each.value.nameservers)
dns_search = [each.value.search_domain]
}]
}]
}) : yamlencode({
version = 1
config = [{
type = "physical"
name = "ens18"
subnets = [{
type = "dhcp"
}]
}]
})
# Instance metadata
meta_data = yamlencode({
instance_id = sha1(each.key)
local-hostname = each.key
})
}

29
terraform/vms.tf
View File

@@ -22,9 +22,22 @@ locals {
# disk_size = "50G"
# }
# Example Test VM with custom git branch (for testing pipeline changes):
# "test-vm" = {
# ip = "10.69.13.100/24"
# flake_branch = "test-pipeline" # Bootstrap from this branch instead of master
# }
# Example Minimal VM using all defaults (uncomment to deploy):
# "minimal-vm" = {}
# "bootstrap-verify-test" = {}
"testvm01" = {
ip = "10.69.13.101/24"
cpu_cores = 2
memory = 2048
disk_size = "20G"
flake_branch = "pipeline-testing-improvements"
}
}
# Compute VM configurations with defaults applied
@@ -44,6 +57,8 @@ locals {
# Network configuration - detect DHCP vs static
ip = lookup(vm, "ip", null)
gateway = lookup(vm, "gateway", var.default_gateway)
# Branch configuration for bootstrap (optional, uses master if not set)
flake_branch = lookup(vm, "flake_branch", null)
}
}
}
@@ -89,8 +104,9 @@ resource "proxmox_vm_qemu" "vm" {
}
ide {
ide2 {
cloudinit {
storage = each.value.storage
# Reference the custom cloud-init disk created in cloud-init.tf
cdrom {
iso = proxmox_cloud_init_disk.ci[each.key].id
}
}
}
@@ -102,15 +118,6 @@ resource "proxmox_vm_qemu" "vm" {
# Agent
agent = 1
# Cloud-init configuration
ciuser = "root"
sshkeys = each.value.ssh_public_key
nameserver = each.value.nameservers
searchdomain = each.value.search_domain
# Network configuration - DHCP or static IP
ipconfig0 = each.value.ip != null ? "ip=${each.value.ip},gw=${each.value.gateway}" : "ip=dhcp"
# Skip IPv6 since we don't use it
skip_ipv6 = true