initial commit

This commit is contained in:
2026-08-27 10:46:36 +02:00
commit 993e5c7327
11 changed files with 591 additions and 0 deletions
+54
View File
@@ -0,0 +1,54 @@
# =============================================================================
# Gitea Actions workflow (active for THIS repository).
# Builds examples/Dockerfile and pushes it to Harbor as a smoke test of the
# k3s runner + DinD + Harbor pipeline.
#
# Required repo/org Actions secrets:
# HARBOR_REGISTRY, HARBOR_USERNAME, HARBOR_PASSWORD
# Optional repo/org Actions variable:
# HARBOR_PROJECT (default: "library")
# =============================================================================
name: build-and-push
on:
push:
branches:
- main
workflow_dispatch:
jobs:
build-and-push:
runs-on: ubuntu-22.04
env:
HARBOR_PROJECT: ${{ vars.HARBOR_PROJECT || 'library' }}
IMAGE_NAME: build-on-k3s-sample
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- name: Compute image tags
id: meta
run: |
IMAGE="${{ secrets.HARBOR_REGISTRY }}/${HARBOR_PROJECT}/${IMAGE_NAME}"
echo "image=${IMAGE}" >> "$GITHUB_OUTPUT"
echo "sha=${GITHUB_SHA::12}" >> "$GITHUB_OUTPUT"
- name: Log in to Harbor
uses: docker/login-action@v3
with:
registry: ${{ secrets.HARBOR_REGISTRY }}
username: ${{ secrets.HARBOR_USERNAME }}
password: ${{ secrets.HARBOR_PASSWORD }}
- name: Build and push
uses: docker/build-push-action@v6
with:
context: ./examples
file: ./examples/Dockerfile
push: true
tags: |
${{ steps.meta.outputs.image }}:latest
${{ steps.meta.outputs.image }}:${{ steps.meta.outputs.sha }}
+8
View File
@@ -0,0 +1,8 @@
# Never commit filled-in secrets — only the *.example templates are tracked.
k3s/10-runner-secret.yaml
k3s/40-harbor-secret.yaml
*.secret.yaml
# Local kube configs
kubeconfig
*.kubeconfig
+137
View File
@@ -0,0 +1,137 @@
# Build Docker images on k3s with Gitea Actions → push to Harbor
This repository contains a complete, reproducible configuration that lets you:
1. Run a **Gitea Actions runner** (`act_runner`) **on a k3s node**.
2. Build **Docker/OCI container images** inside the CI pipeline using a
**Docker-in-Docker (DinD)** sidecar (so plain `docker build` / `docker push` work).
3. **Push the built images to a Harbor** registry.
The whole pipeline is driven by **Gitea Actions** (Gitea's built-in, GitHub-Actions-compatible
CI/CD), so no external CI system (Jenkins/Drone/etc.) is required.
```
┌──────────────┐ registers ┌─────────────────────────────────────┐
│ Gitea │◄────────────────│ act_runner (Deployment on k3s) │
│ (Actions) │ job dispatch │ ┌───────────────┐ ┌──────────────┐ │
│ │────────────────►│ │ act_runner │ │ dockerd │ │
└──────────────┘ │ │ container │─►│ (DinD side- │ │
▲ │ └───────────────┘ │ car, │ │
│ git push │ docker build │ privileged)│ │
┌────┴─────┐ │ docker push └──────────────┘ │
│ developer│ └───────────────────────────┬─────────┘
└──────────┘ │ push image
▼
┌──────────────────┐
│ Harbor registry │
└──────────────────┘
```
---
## Repository layout
| File | Purpose |
|------|---------|
| [k3s/00-namespace.yaml](k3s/00-namespace.yaml) | Dedicated `gitea-runner` namespace. |
| [k3s/10-runner-secret.yaml.example](k3s/10-runner-secret.yaml.example) | Gitea registration token secret (copy & fill in). |
| [k3s/20-runner-config.yaml](k3s/20-runner-config.yaml) | `act_runner` config (`config.yaml`) as a ConfigMap. |
| [k3s/30-runner-deployment.yaml](k3s/30-runner-deployment.yaml) | The `act_runner` + DinD Deployment. |
| [k3s/40-harbor-secret.yaml.example](k3s/40-harbor-secret.yaml.example) | Optional cluster-side Harbor pull secret. |
| [workflows/build-and-push.yaml](workflows/build-and-push.yaml) | Example Gitea Actions workflow (put in your app repo). |
| [examples/Dockerfile](examples/Dockerfile) | Minimal sample image to build. |
| [scripts/deploy.sh](scripts/deploy.sh) | Helper to apply all manifests. |
| [.gitea/workflows/build-and-push.yaml](.gitea/workflows/build-and-push.yaml) | Same workflow, pre-placed so *this* repo self-builds. |
---
## Prerequisites
* A running **k3s** cluster and `kubectl` access to it (`~/.kube/config` or `/etc/rancher/k3s/k3s.yaml`).
* A reachable **Gitea** instance (>= 1.20) with **Actions enabled**
(`[actions] ENABLED = true` in Gitea's `app.ini`).
* A running **Harbor** registry and a **robot account** (recommended) with `push` permission
on the target project.
---
## Step 1 — Enable Gitea Actions (on the Gitea server)
In Gitea's `app.ini`:
```ini
[actions]
ENABLED = true
```
Restart Gitea. Then get a **runner registration token**:
* **Instance level:** *Site Administration → Actions → Runners → Create new Runner* → copy the token.
* **Org/Repo level:** *Settings → Actions → Runners → Create new Runner*.
---
## Step 2 — Create the runner registration secret
```bash
cp k3s/10-runner-secret.yaml.example k3s/10-runner-secret.yaml
# edit k3s/10-runner-secret.yaml and paste your GITEA_INSTANCE_URL + token
kubectl apply -f k3s/10-runner-secret.yaml
```
## Step 3 — Deploy the runner on k3s
```bash
./scripts/deploy.sh
# or manually:
kubectl apply -f k3s/00-namespace.yaml
kubectl apply -f k3s/20-runner-config.yaml
kubectl apply -f k3s/30-runner-deployment.yaml
```
Verify the runner appears **online** in Gitea (*Actions → Runners*) and that both containers are ready:
```bash
kubectl -n gitea-runner rollout status deploy/act-runner
kubectl -n gitea-runner get pods
```
## Step 4 — Configure Harbor credentials for the pipeline
In your **Gitea repository** (or org) → *Settings → Actions → Secrets*, add:
| Secret name | Value |
|-------------|-------|
| `HARBOR_REGISTRY` | e.g. `harbor.example.com` |
| `HARBOR_USERNAME` | Harbor user / robot account, e.g. `robot$ci` |
| `HARBOR_PASSWORD` | the robot account token / password |
> Using a **Harbor robot account** (Project → Robot Accounts) instead of a personal
> login is strongly recommended.
## Step 5 — Add the workflow to your app repo
Copy [workflows/build-and-push.yaml](workflows/build-and-push.yaml) into your application repository at
`.gitea/workflows/build-and-push.yaml`, commit, and push. Gitea will dispatch the job to your
k3s runner, which builds the image via DinD and pushes it to Harbor.
---
## How image building works (no Docker on k3s needed)
k3s uses **containerd**, not Docker, and has no image-build capability. This setup therefore runs a
**dockerd DinD sidecar** next to `act_runner`. The runner container talks to it over
`tcp://localhost:2376` (TLS), so inside jobs the standard `docker` CLI works transparently —
`docker build`, `docker login`, `docker push` all behave like on a normal Docker host.
If you prefer a **rootless / non-privileged** approach (Kaniko or Buildah), see the
"Alternative: rootless builds" section at the bottom of
[workflows/build-and-push.yaml](workflows/build-and-push.yaml).
## Security notes
* The DinD sidecar requires `privileged: true`. Keep this runner in its own namespace and,
ideally, on a dedicated node (see the `nodeSelector`/`tolerations` comments in the Deployment).
* Prefer Harbor **robot accounts** with the least privilege (push to one project only).
* Store all credentials as Kubernetes/Gitea secrets — never commit the filled-in
`*.yaml` (non-`.example`) files. See [.gitignore](.gitignore).
+6
View File
@@ -0,0 +1,6 @@
# Minimal sample image used to smoke-test the build-and-push pipeline.
FROM alpine:3.20
RUN apk add --no-cache ca-certificates
CMD ["sh", "-c", "echo 'Built on k3s via Gitea Actions and pushed to Harbor.'"]
+7
View File
@@ -0,0 +1,7 @@
apiVersion: v1
kind: Namespace
metadata:
name: gitea-runner
labels:
app.kubernetes.io/name: gitea-act-runner
app.kubernetes.io/part-of: ci-cd
+23
View File
@@ -0,0 +1,23 @@
# -----------------------------------------------------------------------------
# Gitea act_runner registration secret.
#
# 1. Copy this file: cp 10-runner-secret.yaml.example 10-runner-secret.yaml
# 2. Fill in the two stringData values below.
# 3. Apply: kubectl apply -f 10-runner-secret.yaml
#
# GITEA_INSTANCE_URL : Base URL of your Gitea server (no trailing /actions).
# GITEA_RUNNER_REGISTRATION_TOKEN :
# Token from Gitea -> (Site Admin | Org | Repo) -> Actions -> Runners
# -> "Create new Runner".
#
# The runner registers itself on first start using these values.
# -----------------------------------------------------------------------------
apiVersion: v1
kind: Secret
metadata:
name: act-runner-registration
namespace: gitea-runner
type: Opaque
stringData:
GITEA_INSTANCE_URL: "https://gitea.example.com"
GITEA_RUNNER_REGISTRATION_TOKEN: "REPLACE_WITH_REGISTRATION_TOKEN"
+38
View File
@@ -0,0 +1,38 @@
# -----------------------------------------------------------------------------
# act_runner configuration, mounted into the runner container as config.yaml.
# Docs: https://docs.gitea.com/usage/actions/act-runner
# -----------------------------------------------------------------------------
apiVersion: v1
kind: ConfigMap
metadata:
name: act-runner-config
namespace: gitea-runner
data:
config.yaml: |
log:
level: info
runner:
# Max parallel jobs this runner will execute.
capacity: 2
timeout: 3h
# Labels advertise which job environments this runner provides.
# The workflow selects one with `runs-on:`.
labels:
- "ubuntu-latest:docker://catthehacker/ubuntu:act-22.04"
- "ubuntu-22.04:docker://catthehacker/ubuntu:act-22.04"
- "dind:docker://catthehacker/ubuntu:act-22.04"
cache:
enabled: true
container:
# Talk to the DinD sidecar over TCP (TLS) instead of a mounted socket.
docker_host: "tcp://localhost:2376"
# Job containers join the default docker network created by dockerd.
network: "host"
# Keep pulled images between jobs to speed builds up.
force_pull: false
# Give job containers access to the DinD TLS certs.
valid_volumes:
- "**"
+153
View File
@@ -0,0 +1,153 @@
# -----------------------------------------------------------------------------
# Gitea act_runner + Docker-in-Docker (DinD) sidecar, running on a k3s node.
#
# - The `dind` container runs a full dockerd (privileged) and exposes it over
# TLS on localhost:2376. Its TLS client certs are shared via an emptyDir.
# - The `runner` container registers with Gitea and dispatches each job to the
# dockerd inside the same pod, so `docker build` / `docker push` "just work".
#
# NOTE: DinD requires a privileged container. Keep this in its own namespace
# and consider pinning it to a dedicated build node (see nodeSelector).
# -----------------------------------------------------------------------------
apiVersion: apps/v1
kind: Deployment
metadata:
name: act-runner
namespace: gitea-runner
labels:
app: act-runner
spec:
# Keep at 1: the registered runner state lives in the runner PVC.
replicas: 1
strategy:
type: Recreate
selector:
matchLabels:
app: act-runner
template:
metadata:
labels:
app: act-runner
spec:
# --- Optionally pin builds to a dedicated node -------------------------
# nodeSelector:
# ci-build: "true" # label a node: kubectl label node <node> ci-build=true
# tolerations:
# - key: "ci-build"
# operator: "Exists"
# effect: "NoSchedule"
# ----------------------------------------------------------------------
containers:
# =====================================================================
# Docker-in-Docker daemon
# =====================================================================
- name: dind
image: docker:27-dind
securityContext:
privileged: true
env:
- name: DOCKER_TLS_CERTDIR
value: /certs
args:
- "--host=tcp://0.0.0.0:2376"
- "--tlsverify"
- "--tlscacert=/certs/ca/cert.pem"
- "--tlscert=/certs/server/cert.pem"
- "--tlskey=/certs/server/key.pem"
volumeMounts:
- name: docker-certs
mountPath: /certs
- name: docker-storage
mountPath: /var/lib/docker
readinessProbe:
exec:
command: ["docker", "-H", "tcp://localhost:2376", "--tlsverify",
"--tlscacert=/certs/ca/cert.pem",
"--tlscert=/certs/client/cert.pem",
"--tlskey=/certs/client/key.pem", "info"]
initialDelaySeconds: 15
periodSeconds: 10
resources:
requests:
cpu: "250m"
memory: "512Mi"
limits:
cpu: "2"
memory: "4Gi"
# =====================================================================
# Gitea Actions runner
# =====================================================================
- name: runner
image: gitea/act_runner:0.2.11
env:
# Registration data (from the Secret).
- name: GITEA_INSTANCE_URL
valueFrom:
secretKeyRef:
name: act-runner-registration
key: GITEA_INSTANCE_URL
- name: GITEA_RUNNER_REGISTRATION_TOKEN
valueFrom:
secretKeyRef:
name: act-runner-registration
key: GITEA_RUNNER_REGISTRATION_TOKEN
- name: GITEA_RUNNER_NAME
valueFrom:
fieldRef:
fieldPath: metadata.name
# Point the runner at the DinD daemon over TLS.
- name: DOCKER_HOST
value: "tcp://localhost:2376"
- name: DOCKER_CERT_PATH
value: "/certs/client"
- name: DOCKER_TLS_VERIFY
value: "1"
# Path to the mounted act_runner config.
- name: CONFIG_FILE
value: "/config/config.yaml"
volumeMounts:
- name: docker-certs
mountPath: /certs
readOnly: true
- name: runner-config
mountPath: /config
readOnly: true
- name: runner-data
mountPath: /data
resources:
requests:
cpu: "100m"
memory: "128Mi"
limits:
cpu: "1"
memory: "1Gi"
volumes:
# Shared TLS certs between dockerd and the runner/job containers.
- name: docker-certs
emptyDir: {}
# dockerd image/layer storage (ephemeral; use a PVC for a persistent cache).
- name: docker-storage
emptyDir: {}
- name: runner-config
configMap:
name: act-runner-config
# Persists the runner registration (.runner) across restarts.
- name: runner-data
persistentVolumeClaim:
claimName: act-runner-data
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: act-runner-data
namespace: gitea-runner
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
# k3s default storage class (local-path). Change if you use another provisioner.
# storageClassName: local-path
+29
View File
@@ -0,0 +1,29 @@
# -----------------------------------------------------------------------------
# OPTIONAL: cluster-side Harbor pull secret.
#
# This is NOT needed for the CI build/push (that uses Gitea Actions secrets).
# Create it only if the same k3s cluster must *pull* the private images that the
# pipeline pushes to Harbor (e.g. when you later deploy those images here).
#
# Easiest way to create it (recommended over editing this file by hand):
#
# kubectl create secret docker-registry harbor-pull \
# --namespace <your-app-namespace> \
# --docker-server=harbor.example.com \
# --docker-username='robot$ci' \
# --docker-password='REPLACE_WITH_ROBOT_TOKEN'
#
# Then reference it from your workloads:
# spec.imagePullSecrets: [{ name: harbor-pull }]
#
# The manifest below is the equivalent declarative form (base64 dockerconfigjson).
# -----------------------------------------------------------------------------
apiVersion: v1
kind: Secret
metadata:
name: harbor-pull
namespace: default
type: kubernetes.io/dockerconfigjson
data:
# echo -n '{"auths":{"harbor.example.com":{"username":"robot$ci","password":"TOKEN","auth":"<base64 user:pass>"}}}' | base64 -w0
.dockerconfigjson: REPLACE_WITH_BASE64_DOCKERCONFIGJSON
+39
View File
@@ -0,0 +1,39 @@
#!/usr/bin/env bash
# Deploy the Gitea act_runner (with DinD) onto k3s.
#
# Usage:
# ./scripts/deploy.sh
#
# Prerequisites:
# - kubectl configured against your k3s cluster
# (e.g. export KUBECONFIG=/etc/rancher/k3s/k3s.yaml)
# - k3s/10-runner-secret.yaml created from the .example and filled in.
set -euo pipefail
ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
K3S_DIR="${ROOT_DIR}/k3s"
if [[ ! -f "${K3S_DIR}/10-runner-secret.yaml" ]]; then
echo "ERROR: ${K3S_DIR}/10-runner-secret.yaml not found." >&2
echo " cp ${K3S_DIR}/10-runner-secret.yaml.example ${K3S_DIR}/10-runner-secret.yaml" >&2
echo " then fill in GITEA_INSTANCE_URL and the registration token." >&2
exit 1
fi
echo ">> Applying namespace"
kubectl apply -f "${K3S_DIR}/00-namespace.yaml"
echo ">> Applying runner registration secret"
kubectl apply -f "${K3S_DIR}/10-runner-secret.yaml"
echo ">> Applying runner config"
kubectl apply -f "${K3S_DIR}/20-runner-config.yaml"
echo ">> Applying runner deployment"
kubectl apply -f "${K3S_DIR}/30-runner-deployment.yaml"
echo ">> Waiting for rollout"
kubectl -n gitea-runner rollout status deploy/act-runner --timeout=180s
echo ">> Done. Runner pods:"
kubectl -n gitea-runner get pods -o wide
+97
View File
@@ -0,0 +1,97 @@
# =============================================================================
# Gitea Actions workflow: build a Docker image and push it to Harbor.
#
# Place this file in your application repository at:
# .gitea/workflows/build-and-push.yaml
#
# Required repo/org Actions secrets (Settings -> Actions -> Secrets):
# HARBOR_REGISTRY e.g. harbor.example.com
# HARBOR_USERNAME e.g. robot$ci
# HARBOR_PASSWORD the robot account token
#
# Optional repo/org Actions variables (Settings -> Actions -> Variables):
# HARBOR_PROJECT Harbor project name (default: "library")
# IMAGE_NAME image name (default: repository name)
# =============================================================================
name: build-and-push
on:
push:
branches:
- main
tags:
- "v*"
workflow_dispatch:
jobs:
build-and-push:
# Uses the label advertised by the act_runner on k3s.
runs-on: ubuntu-22.04
env:
HARBOR_PROJECT: ${{ vars.HARBOR_PROJECT || 'library' }}
IMAGE_NAME: ${{ vars.IMAGE_NAME || gitea.event.repository.name }}
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- name: Compute image tags
id: meta
run: |
REGISTRY="${{ secrets.HARBOR_REGISTRY }}"
IMAGE="${REGISTRY}/${HARBOR_PROJECT}/${IMAGE_NAME}"
SHA_TAG="${GITHUB_SHA::12}"
echo "image=${IMAGE}" >> "$GITHUB_OUTPUT"
echo "sha_tag=${SHA_TAG}" >> "$GITHUB_OUTPUT"
# Use the git tag as the version when the ref is a tag, else 'latest'.
if [[ "${GITHUB_REF}" == refs/tags/* ]]; then
echo "version=${GITHUB_REF_NAME}" >> "$GITHUB_OUTPUT"
else
echo "version=latest" >> "$GITHUB_OUTPUT"
fi
- name: Log in to Harbor
uses: docker/login-action@v3
with:
registry: ${{ secrets.HARBOR_REGISTRY }}
username: ${{ secrets.HARBOR_USERNAME }}
password: ${{ secrets.HARBOR_PASSWORD }}
- name: Build and push
uses: docker/build-push-action@v6
with:
context: .
file: ./Dockerfile
push: true
tags: |
${{ steps.meta.outputs.image }}:${{ steps.meta.outputs.version }}
${{ steps.meta.outputs.image }}:${{ steps.meta.outputs.sha_tag }}
- name: Summary
run: |
echo "Pushed:"
echo " ${{ steps.meta.outputs.image }}:${{ steps.meta.outputs.version }}"
echo " ${{ steps.meta.outputs.image }}:${{ steps.meta.outputs.sha_tag }}"
# =============================================================================
# Alternative: rootless builds (no privileged DinD sidecar)
# -----------------------------------------------------------------------------
# If you cannot run a privileged DinD sidecar, build with Kaniko instead.
# Replace the "Build and push" step (and drop the Buildx/login steps) with a
# container step running the Kaniko executor. Kaniko logs in to Harbor via a
# generated docker config:
#
# - name: Build and push with Kaniko
# uses: https://github.com/int128/kaniko-action@v1
# with:
# push: true
# tags: ${{ steps.meta.outputs.image }}:${{ steps.meta.outputs.version }}
# registry: ${{ secrets.HARBOR_REGISTRY }}
# username: ${{ secrets.HARBOR_USERNAME }}
# password: ${{ secrets.HARBOR_PASSWORD }}
#
# In that case the act_runner does not need the DinD sidecar; the Kaniko
# executor image performs the build inside the job container.
# =============================================================================