Prerequisites and Considerations

To deploy FME Flow with Kubernetes, you must fulfill the following prerequisites:

Install Kubernetes

Install and configure a Kubernetes cluster. To install a local (single-node) cluster, we recommend installing Docker Desktop, and following the Getting Started guide. Once installed, select Enable Kubernetes to start a single-node cluster.

Tip Ensure sufficient CPU and memory resources are assigned to your Docker Engine, preferably higher than the default allocations. We recommend at least 4 CPUs and 8 GB memory.

Install Helm

The Helm tool is required for pulling the fmeflow Kubernetes chart.

Install Helm.
Add the fmeflow chart to the cluster:

helm repo add safesoftware https://safesoftware.github.io/helm-charts/

Install NGINX Ingress Controller

FME Flow deployments require the official NGINX Ingress Controller from NGINX Inc. Install it using Helm:

helm install nginx-ingress oci://ghcr.io/nginx/charts/nginx-ingress

For full installation options and prerequisites, see the NGINX Ingress Controller Helm installation guide.

Other Considerations

When deploying FME Flow to a Kubernetes cluster, keep in mind the following:

It is not possible to use the FME Flow Web User Interface to deploy FME Engines or change the number of FME Engines. For more information, see Defining FME Engines, Queue Control, and Streams to a Kubernetes Deployment.
Any memory or CPU limits you define on containers are ignored by all FME Flow processes. Memory and CPU limits defined on FME Engine containers are honored.
Pods for the FME Engines and FME Flow Core/Web Services can be scaled; no other Pods can be scaled.