Kubernetes Deployment

Prerequisites

Requires Helm (https://helm.sh/) and an active connection to a kubernetes cluster with a running tiller server.

Tested with:

• K3s v1.18.8+k3s1 (6b595318) with K8s v1.18.8
• Helm v3.1.2

Usage

We provide two helm chart options to get you going:

• default: a light-weight option with few pipeline elements, needs less memory
• full: contains more pipeline elements, requires >16 GB RAM (recommended)

Starting the default helm chart option is as easy as simply running the following command from the root of this folder:

NOTE: Starting might take a while since we also initially pull all Docker images from Dockerhub.

helm install streampipes ./

After a while, all containers should successfully started, indicated by the Running status.

kubectl get pods
NAME                                           READY   STATUS    RESTARTS   AGE
activemq-66d58f47cf-2r2nb                      1/1     Running   0          3m27s
backend-76ddc486c8-nswpc                       1/1     Running   0          3m27s
connect-master-7b477f9b79-8dfvr                1/1     Running   0          3m26s
connect-worker-78d89c989c-9v8zs                1/1     Running   0          3m27s
consul-55965f966b-gwb7l                        1/1     Running   0          3m27s
couchdb-77db98cf7b-xnnvb                       1/1     Running   0          3m27s
influxdb-b95b6479-r8wh8                        1/1     Running   0          3m27s
kafka-657b5fb77-dp2d6                          1/1     Running   0          3m27s
pipeline-elements-all-jvm-79c445dbd9-m8xcs     1/1     Running   0          3m27s
sources-watertank-simulator-6c6b8844f6-6b4d7   1/1     Running   0          3m27s
ui-b94bd9766-rm6zb                             2/2     Running   0          3m27s
zookeeper-5d9947686f-6nzgs                     1/1     Running   0          3m26s

After all containers are successfully started just got to your browser and visit any of the k8s cluster nodes on http://<NODE_IP> to finish the installation.

NOTE: If you're running Docker for Mac or Docker for Windows with a local k8s cluster, the above step to use your host IP might not work. Luckily, you can port-forward a service port to your localhost using the following command to be able to access the UI either via http://localhost or http://<HOST_IP> (you require sudo to run this command in order to bind to a privileged port).

kubectl port-forward svc/ui --address=0.0.0.0 80:80

Starting the full helm chart option is almost the same:

helm install streampipes ./ --set deployment=full

Deleting the current helm chart deployment:

helm del streampipes