FoundationDB is a distributed database designed to handle large volumes of structured data across clusters of commodity servers. It organizes data as an ordered key-value store and employs ACID transactions for all operations. It is especially well-suited for read/write workloads but also has excellent performance for write-intensive workloads. Users interact with the database using API language binding.

To learn more about FoundationDB, visit foundationdb.org

Overview

This project provides an operator for managing FoundationDB clusters on Kubernetes. Some more details are covered in this YouTube video: Operating FoundationDB on Kubernetes (DoK Day EU 2022).

Running the Operator

To run the operator in your environment, you need to install the operator and the CRDs: Note this will install the latest version from main. For a production setup you should refer to a specific tag.

kubectl apply -f https://raw.githubusercontent.com/FoundationDB/fdb-kubernetes-operator/main/config/crd/bases/apps.foundationdb.org_foundationdbclusters.yaml
kubectl apply -f https://raw.githubusercontent.com/FoundationDB/fdb-kubernetes-operator/main/config/crd/bases/apps.foundationdb.org_foundationdbbackups.yaml
kubectl apply -f https://raw.githubusercontent.com/FoundationDB/fdb-kubernetes-operator/main/config/crd/bases/apps.foundationdb.org_foundationdbrestores.yaml
kubectl apply -f https://raw.githubusercontent.com/foundationdb/fdb-kubernetes-operator/main/config/samples/deployment.yaml

At this point, you can set up a sample cluster:

kubectl apply -f https://raw.githubusercontent.com/foundationdb/fdb-kubernetes-operator/main/config/samples/cluster.yaml

You can see logs from the operator by running kubectl logs -f -l app=fdb-kubernetes-operator-controller-manager --container=manager. To determine whether the reconciliation has completed, you can run kubectl get foundationdbcluster test-cluster. This will show the latest generation of the spec and the last reconciled generation of the spec. Once reconciliation has completed, these values will be the same.

Once the reconciliation is complete, you can run kubectl fdb exec -it test-cluster -- fdbcli to open up a CLI on your cluster.

You can also browse the sample directory for more examples of different resource configurations.

For more information about using the operator, including detailed discussion of how to customize your deployments, see the user manual.

For more information on version compatibility, see our compatibility guide.

For more information on the fields you can define on the cluster resource, see the API documentation.

Using helm

This repository contains a helm chart under charts. The helm chart is not published and not actively tested by our CI. The charts are provided by the community on a best-effort base.

Local Development

Environment Set-up

1. Install Go on your machine, see the Getting Started guide for more information.
2. Install the required dependencies with make deps.
3. Install the foundationDB client package, if you're using an arm64 Mac, make sure you install the arm64 package.

Running Locally

To get this controller running in a local Kubernetes cluster:

1. The assumption is that you have a local Kubernetes cluster running. Depending on what solution you use some of the following steps might differ.
2. Clone this repository onto your local machine.
3. Run config/test-certs/generate_secrets.bash to set up a secret with self-signed test certs.
4. Run make rebuild-operator to install the operator. By default, the container image is built for the platform where this command is executed. To override the platform, for example, to build an amd64 image on Apple M1, you can set the BUILD_PLATFORM env variable BUILD_PLATFORM="linux/amd64" make rebuild-operator.
5. Run kubectl apply -k ./config/tests/base to create a new FoundationDB cluster with the operator.

NOTE: FoundationDB currently only publishes container images for running on amd64/x64 nodes.

Running Locally with nerdctl

Instead of Docker you can also use nerdctl to build and push your images. In order to use a different image builder than docker you can use the env variable BUILDER:

# This will use nerdctl for building the image in the k8s.io namespace
export BUILDER='nerdctl -n k8s.io'

You can test your setup with SKIP_TEST=1 make container-build which will build the image locally. After the command successfully finished you can verify with nerdctl -n k8s.io images fdb-kubernetes-operator:latest that the image is available.

Running e2e tests with the operator

This repository contains e2e tests to verify that the operator behaves correct in different situations. Besides functional tests the e2e test suite uses chaos-mesh to inject failures during test runs. The test suite is build in a way that you can run the e2e tests against different Kubernetes clusters.

Customizing Your Build

The makefile supports environment variables that allow you to customize your build. You can use these to push to custom docker repos and deployment platforms.

• IMG: This specifies the image that gets built for the operator.
• SIDECAR_IMG: This specifies the image for the foundationdb-kubernetes-sidecar process used in init containers for the operator. This does not change the images used for the FoundationDB clusters, which are specified in the cluster spec.
• REMOTE_BUILD: This can be set to 1 to indicate that you are running the operator in a remote environment, rather than on your local development machine. This will activate a remote build patch, which changes the image pull policy in the operator's pod spec. Setting this also tells the Makefile to push images as part of the rebuild-operator command.
• FDB_WEBSITE: This specifies the base path for the website used to download FDB client packages in the docker builds. You can use this to download custom binaries from your own host, provided that your path structure matches the paths expected in the Dockerfile.

Known Limitations

1. Support for backups in the operator is still in development, and there are significant missing features.
2. Additional limitations can be found under Warnings.
3. The created FoundationDB cluster is only reachable from within the Kubernetes cluster. Except if the assigned IP addresses for the Pods are also reachable outside of the Kubernetes cluster.