PlexStack Part 4 – Our first app: Tautulli

We are now at a point where we can build our first application that requires some persistence. We are going to start with Tautulli, an application that provides statistics about your Plex server.

We assume that you only have a single server. The state of kubernetes storage is interesting. The easiest way is to simply pass a host path in to the pod, but that doesn’t work when you have multiple nodes. Incidentally, what I do for my day job (Portworx Cloud Architect) is solving these problems for customers. More on that later.

We first need to specify a location to store configuration data. I will use /opt/plexstack/tautulli as an example.

mkdir -p /opt/plexstack/tautulli

Shell
​x
 
mkdir -p /opt/plexstack/tautulli

Next, let’s take a look at the manifest to install tautulli:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: tautulli-deployment
  labels:
    app: tautulli
spec:
  replicas: 1
  selector:
     matchLabels:
       app: tautulli
  template:
    metadata:
      labels:
        app: tautulli
    spec:
     containers:
        - name: tautulli
          image: ghcr.io/linuxserver/tautulli
          env:
            - name: PUID
              value: "999"
            - name: PGID
              value: "999"
            - name: TZ
              value: "America/Los_Angeles"
          ports:
            - containerPort: 8181
          volumeMounts:
            - mountPath: /config
              name: tautulli-config
     volumes:
       - name: tautulli-config
         hostPath: 
            path: /opt/plexstack/tautulli
---
kind: Service
apiVersion: v1
metadata:
  name: tautulli-service
spec:
  selector:
    app: tautulli
  ports:
  - protocol: TCP
    port: 8181
    targetPort: 8181
  type: LoadBalancer
---
kind: Ingress
apiVersion: networking.k8s.io/v1
metadata:
  name: ingress-tautulli
  namespace: plexstack
  annotations:
    cert-manager.io/cluster-issuer: letsencrypt-prod
    kubernetes.io/ingress.class: nginx
spec:
  tls:
    - hosts:
        - tautulli.ccrow.org
      secretName: tautulli-tls
  rules:
    - host: tautulli.ccrow.org
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: tautulli-service
                port:
                  number: 8181

tautulli.yaml
 
apiVersion: apps/v1
kind: Deployment
metadata:
  name: tautulli-deployment
  labels:
    app: tautulli
spec:
  replicas: 1
  selector:
     matchLabels:
       app: tautulli
  template:
    metadata:
      labels:
        app: tautulli
    spec:
     containers:
        - name: tautulli
          image: ghcr.io/linuxserver/tautulli
          env:
            - name: PUID
              value: "999"
            - name: PGID
              value: "999"
            - name: TZ
              value: "America/Los_Angeles"
          ports:
            - containerPort: 8181
          volumeMounts:
            - mountPath: /config
              name: tautulli-config
     volumes:
       - name: tautulli-config
         hostPath: 
            path: /opt/plexstack/tautulli
---
kind: Service
apiVersion: v1
metadata:
  name: tautulli-service
spec:
  selector:
    app: tautulli
  ports:
  - protocol: TCP
    port: 8181
    targetPort: 8181
  type: LoadBalancer
---
kind: Ingress
apiVersion: networking.k8s.io/v1
metadata:
  name: ingress-tautulli
  namespace: plexstack
  annotations:
    cert-manager.io/cluster-issuer: letsencrypt-prod
    kubernetes.io/ingress.class: nginx
spec:
  tls:
    - hosts:
        - tautulli.ccrow.org
      secretName: tautulli-tls
  rules:
    - host: tautulli.ccrow.org
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: tautulli-service
                port:
                  number: 8181

There is a lot to unpack here:

The first section is the deployment. It defines the application that will run. Line 19 specifies the image.
Lines 21 – 26 are environment variables that configure tautulli
We can see where we specify the /config directory inside the container to be mapped to a host path (lines 29 – 35).
The next section is the service, which looks for pods with an app selector of tautulli.
We are also going to provision a load balancer IP address to help with troubleshooting. This could be changed to ClusterIP to be internal only. After all, why go to an ip address when we can use an ingress.
Tautulli.ccrow.org must resolve to our rancher node through the firewall (a step we already did in the last blog.

Let’s apply the manifest with:

# create the namespace
kubectl create namespace plexstack

# apply the manifest
kubectl -n plexstack apply -f tautulli.yaml

# check on the deployment
kubectl -n plexstack get all -o wide
NAME                                      READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
pod/tautulli-deployment-b4d5485df-f28px   1/1     Running   0          45s   10.42.2.30   rke04   <none>           <none>

NAME                       TYPE           CLUSTER-IP   EXTERNAL-IP   PORT(S)          AGE   SELECTOR
service/tautulli-service   LoadBalancer   10.43.36.8   10.0.1.55     8181:31154/TCP   45s   app=tautulli

NAME                                  READY   UP-TO-DATE   AVAILABLE   AGE   CONTAINERS   IMAGES                         SELECTOR
deployment.apps/tautulli-deployment   1/1     1            1           45s   tautulli     ghcr.io/linuxserver/tautulli   app=tautulli

NAME                                            DESIRED   CURRENT   READY   AGE   CONTAINERS   IMAGES                         SELECTOR
replicaset.apps/tautulli-deployment-b4d5485df   1         1         1       45s   tautulli     ghcr.io/linuxserver/tautulli   app=tautulli,pod-template-hash=b4d5485df

Shell
 
# create the namespace
kubectl create namespace plexstack
​
# apply the manifest
kubectl -n plexstack apply -f tautulli.yaml
​
# check on the deployment
kubectl -n plexstack get all -o wide
NAME                                      READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
pod/tautulli-deployment-b4d5485df-f28px   1/1     Running   0          45s   10.42.2.30   rke04   <none>           <none>
​
NAME                       TYPE           CLUSTER-IP   EXTERNAL-IP   PORT(S)          AGE   SELECTOR
service/tautulli-service   LoadBalancer   10.43.36.8   10.0.1.55     8181:31154/TCP   45s   app=tautulli
​
NAME                                  READY   UP-TO-DATE   AVAILABLE   AGE   CONTAINERS   IMAGES                         SELECTOR
deployment.apps/tautulli-deployment   1/1     1            1           45s   tautulli     ghcr.io/linuxserver/tautulli   app=tautulli
​
NAME                                            DESIRED   CURRENT   READY   AGE   CONTAINERS   IMAGES                         SELECTOR
replicaset.apps/tautulli-deployment-b4d5485df   1         1         1       45s   tautulli     ghcr.io/linuxserver/tautulli   app=tautulli,pod-template-hash=b4d5485df

Notice the external IP address that was created for the tautulli-service. You can connect to the app from that IP (be sure to add the 8181 port!) instead of the DNS name.

All configuration data will be stored under /opt/plexstack/tautulli on your node.

Bonus Appplication: smtp

In order for tautulli to send email, we need to set up an SMTP server. This will really show off the power of kubernetes configurations. Take a look at this manifest:

---
 apiVersion: apps/v1
 kind: Deployment
 metadata:
   name: smtp-deployment
   labels:
     app: smtp
 spec:
   replicas: 1
   selector:
      matchLabels:
        app: smtp
   template:
     metadata:
       labels:
         app: smtp

     spec:
      containers:
         - name: smtp
           image: pure/smtp-relay
           env:
             - name: SMTP_HOSTNAME
               value: "mail.ccrow.org"
             - name: RELAY_NETWORKS
               value: "10.0.0.0/8"
           ports:
             - containerPort: 25
---
kind: Service
apiVersion: v1
metadata:
  name: smtp-service
spec:
  selector:
    app: smtp
  ports:
  - protocol: TCP
    port: 25
    targetPort: 25
  type: ClusterIP

smtp.yaml
 
---
 apiVersion: apps/v1
 kind: Deployment
 metadata:
   name: smtp-deployment
   labels:
     app: smtp
 spec:
   replicas: 1
   selector:
      matchLabels:
        app: smtp
   template:
     metadata:
       labels:
         app: smtp
​
     spec:
      containers:
         - name: smtp
           image: pure/smtp-relay
           env:
             - name: SMTP_HOSTNAME
               value: "mail.ccrow.org"
             - name: RELAY_NETWORKS
               value: "10.0.0.0/8"
           ports:
             - containerPort: 25
---
kind: Service
apiVersion: v1
metadata:
  name: smtp-service
spec:
  selector:
    app: smtp
  ports:
  - protocol: TCP
    port: 25
    targetPort: 25
  type: ClusterIP

You can apply the above manifest. Be sure to change lines 24 and 26 to match your network. Please note: “Your network” really means your internal kubernetes network. After all, why would we send an email from an external source (well, unless you want to, in which case, change line 41 to loadBalancer).

kubectl -n plexstack apply -f smtp.yaml

Shell
 
kubectl -n plexstack apply -f smtp.yaml

We now have a working SMTP server! The coolest part of kubernetes service discovery is being able to simply use the name of our service for any application in the same namespace:

Using the service name means that this configuration is portable, no need to actually plug in the cluster IP address that was assigned.