B23 Kubernetes Operator for NiFi-Stateless

An Operator for scheduling and executing NiFi Flows on Kubernetes. The operator is made possible by NiFi-Stateless

This is a proof of concept for the proposed Kubernetes Runtime for nifi-stateless

Install the Operator on a cluster

If you just want to run the operator on your cluster:

kubectl apply -f https://raw.githubusercontent.com/b23llc/nifi-stateless-operator/master/config/deploy/nifi-stateless-operator.yaml

If you also want to run a Nifi/Registry Pod to use as a canvas as a convenience for developing and testing flows, run:

kubectl apply -f https://raw.githubusercontent.com/b23llc/nifi-stateless-operator/master/config/deploy/nifi.yaml

To access the nifi/registry services, run: kubectl -n nifi-stateless-operator-system port-forward statefulset/nifi 8081:8081 18080:18080 Note that this nifi/registry pod is a convenience and should not be used for production workloads.

Test the operator by creating a NiFiFn Resource. Make sure to update the flow and bucket fields with the uuid of a flow from your registry.

# config/samples/nififn_v1alpha1_nififn.yaml
apiVersion: nififn.nifi-stateless.b23.io/v1alpha1
kind: NiFiFn
metadata:
  name: nififn-sample
spec:
  # image: "dbkegley/nifi-stateless:1.10.0-SNAPSHOT"
  runFrom: registry
  registryUrl: "http://registry-service:18080"
  bucketId: "8444dc91-00f3-415c-a965-256ffa28c3f5"
  flowId: "d6045598-3d11-438d-b921-52d466b66314"
  flowVersion: -1
  flowFiles:
  - absolute.path: /path/to/input/data/
    filename: testfile.txt
  - absolute.path: /path/to/input/data/
    filename: hello.txt
    nifi_content: "hello world"

It is also possible to run from a flow.xml.gz file:

FROM dbkegley/nifi-stateless:1.10.0-SNAPSHOT
ADD flow.xml.gz /opt/nifi/nifi-current/conf/flow.xml.gz

and update the NiFiFn spec to set: runFrom: xml

Build

Requires:

Develop

Install CRDs: make install

Run Operator: make run

Release

Run make gen-release target to test/build/tag the docker image, generate the CRDs, generate operator install and rbac manifests and push the image to docker hub

Tested locally with docker-for-mac Version 2.0.0.2 (30215) and on Google Cloud Platform with Google Kubernetes Engine

User notes