9.9 KiB
Kanboard
Kanboard is a Kanban tool, developed by Frédéric Guillot. (Who also happens to be the developer of my favorite RSS reader, Miniflux)
Features include:
- Visualize your work
- Limit your work in progress to be more efficient
- Customize your boards according to your business activities
- Multiple projects with the ability to drag and drop tasks
- Reports and analytics
- Fast and simple to use
- Access from anywhere with a modern browser
- Plugins and integrations with external services
- Free, open source and self-hosted
- Super simple installation
Ingredients
- A Kubernetes Cluster including Traefik Ingress
- A DNS name for your kanboard instance (kanboard.example.com, below) pointing to your load balancer, fronting your Traefik ingress
Preparation
Prepare traefik for namespace
When you deployed Traefik via the helm chart, you would have customized values.yml for your deployment. In values.yml is a list of namespaces which Traefik is permitted to access. Update values.yml to include the kanboard namespace, as illustrated below:
<snip>
kubernetes:
namespaces:
- kube-system
- nextcloud
- kanboard
- miniflux
<snip>
If you've updated values.yml, upgrade your traefik deployment via helm, by running helm upgrade --values values.yml traefik stable/traefik --recreate-pods
Create data locations
Although we could simply bind-mount local volumes to a local Kubuernetes cluster, since we're targetting a cloud-based Kubernetes deployment, we only need a local path to store the YAML files which define the various aspects of our Kubernetes deployment.
mkdir /var/data/config/kanboard
Create namespace
We use Kubernetes namespaces for service discovery and isolation between our stacks, so create a namespace for the kanboard stack with the following .yml:
cat <<EOF > /var/data/config/kanboard/namespace.yml
apiVersion: v1
kind: Namespace
metadata:
name: kanboard
EOF
kubectl create -f /var/data/config/kanboard/namespace.yaml
Create persistent volume claim
Persistent volume claims are a streamlined way to create a persistent volume and assign it to a container in a pod. Create a claim for the kanboard app and plugin data:
cat <<EOF > /var/data/config/kanboard/persistent-volumeclaim.yml
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: kanboard-volumeclaim
namespace: kanboard
annotations:
backup.kubernetes.io/deltas: P1D P7D
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 1Gi
EOF
kubectl create -f /var/data/config/kanboard/kanboard-volumeclaim.yaml
!!! question "What's that annotation about?" The annotation is used by k8s-snapshots to create daily incremental snapshots of your persistent volumes. In this case, our volume is snapshotted daily, and copies kept for 7 days.
Create ConfigMap
Kanboard's configuration is all contained within config.php, which needs to be presented to the container. We could maintain config.php in the persistent volume we created above, but this would require manually accessing the pod every time we wanted to make a change.
Instead, we'll create config.php as a ConfigMap, meaning it "lives" within the Kuberetes cluster and can be presented to our pod. When we want to make changes, we simply update the ConfigMap (delete and recreate, to be accurate), and relaunch the pod.
Grab a copy of config.default.php, save it to /var/data/config/kanboard/config.php, and customize it per the guide.
At the very least, I'd suggest making the following changes:
define('PLUGIN_INSTALLER', true); // Yes, I want to install plugins using the UI
define('ENABLE_URL_REWRITE', false); // Yes, I want pretty URLs
Now create the configmap from config.php, by running kubectl create configmap -n kanboard kanboard-config --from-file=config.php
Serving
Now that we have a namespace, a persistent volume, and a configmap, we can create a deployment, service, and ingress for the kanboard pod.
Create deployment
Create a deployment to tell Kubernetes about the desired state of the pod (which it will then attempt to maintain). Note below that we mount the persistent volume twice, to both /var/www/app/data and /var/www/app/plugins, using the subPath value to differentiate them. This trick avoids us having to provision two persistent volumes just for data mounted in 2 separate locations.
--8<-- "premix-cta.md"
cat <<EOF > /var/data/kanboard/deployment.yml
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
namespace: kanboard
name: app
labels:
app: app
spec:
replicas: 1
selector:
matchLabels:
app: app
template:
metadata:
labels:
app: app
spec:
containers:
- image: kanboard/kanboard
name: app
volumeMounts:
- name: kanboard-config
mountPath: /var/www/app/config.php
subPath: config.php
- name: kanboard-app
mountPath: /var/www/app/data
subPath: data
- name: kanboard-app
mountPath: /var/www/app/plugins
subPath: plugins
volumes:
- name: kanboard-app
persistentVolumeClaim:
claimName: kanboard-app
- name: kanboard-config
configMap:
name: kanboard-config
EOF
kubectl create -f /var/data/kanboard/deployment.yml
Check that your deployment is running, with kubectl get pods -n kanboard. After a minute or so, you should see a "Running" pod, as illustrated below:
[funkypenguin:~] % kubectl get pods -n kanboard
NAME READY STATUS RESTARTS AGE
app-79f97f7db6-hsmfg 1/1 Running 0 11d
[funkypenguin:~] %
Create service
The service resource "advertises" the availability of TCP port 80 in your pod, to the rest of the cluster (constrained within your namespace). It seems a little like overkill coming from the Docker Swarm's automated "service discovery" model, but the Kubernetes design allows for load balancing, rolling upgrades, and health checks of individual pods, without impacting the rest of the cluster elements.
cat <<EOF > /var/data/kanboard/service.yml
kind: Service
apiVersion: v1
metadata:
name: app
namespace: kanboard
spec:
selector:
app: app
ports:
- protocol: TCP
port: 80
clusterIP: None
EOF
kubectl create -f /var/data/kanboard/service.yml
Check that your service is deployed, with kubectl get services -n kanboard. You should see something like this:
[funkypenguin:~] % kubectl get service -n kanboard
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
app ClusterIP None <none> 80/TCP 38d
[funkypenguin:~] %
Create ingress
The ingress resource tells Traefik what to forward inbound requests for kanboard.example.com to your service (defined above), which in turn passes the request to the "app" pod. Adjust the config below for your domain.
cat <<EOF > /var/data/kanboard/ingress.yml
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
name: app
namespace: kanboard
annotations:
kubernetes.io/ingress.class: traefik
spec:
rules:
- host: kanboard.example.com
http:
paths:
- backend:
serviceName: app
servicePort: 80
EOF
kubectl create -f /var/data/kanboard/ingress.yml
Check that your service is deployed, with kubectl get ingress -n kanboard. You should see something like this:
[funkypenguin:~] % kubectl get ingress -n kanboard
NAME HOSTS ADDRESS PORTS AGE
app kanboard.funkypenguin.co.nz 80 38d
[funkypenguin:~] %
Access Kanboard
At this point, you should be able to access your instance on your chosen DNS name (i.e. https://kanboard.example.com)
Updating config.php
Since config.php is a ConfigMap now, to update it, make your local changes, and then delete and recreate the ConfigMap, by running:
kubectl delete configmap -n kanboard kanboard-config
kubectl create configmap -n kanboard kanboard-config --from-file=config.php
Then, in the absense of any other changes to the deployement definition, force the pod to restart by issuing a "null patch", as follows:
kubectl patch -n kanboard deployment app -p "{\"spec\":{\"template\":{\"metadata\":{\"labels\":{\"date\":\"`date +'%s'`\"}}}}}"
Troubleshooting
To look at the Kanboard pod's logs, run kubectl logs -n kanboard <name of pod per above> -f. For further troubleshooting hints, see Troubleshooting.
--8<-- "recipe-footer.md"