Quick start v1

Suggest edits

You can test an EDB Postgres Distributed (PGD) cluster on your laptop or computer using EDB Postgres Distributed for Kubernetes on a single local Kubernetes cluster built with Kind.

Warning

These instructions are only for demonstration, testing, and practice purposes and must not be used in production.

This quick start shows you how to start an EDB Postgres Distributed cluster on your local Kubernetes installation so you can experiment with it.

Important

To connect to the Kubernetes cluster, make sure that you have kubectl installed on your machine. See the Kubernetes documentation on installing kubectl.

Part 1 - Set up the local Kubernetes playground

Install Kind, a tool for running local Kubernetes clusters using Docker container nodes. (Kind stands for Kubernetes IN Docker.) If you already have access to a Kubernetes cluster, you can skip to Part 2.

Install Kind on your environment following the instructions in Kind Quick Start. Then, create a Kubernetes cluster:

kind create cluster --name pgd

Part 2 - Install EDB Postgres Distributed for Kubernetes

After you have a Kubernetes installation up and running on your laptop, you can install EDB Postgres Distributed for Kubernetes.

See Installation for details.

Part 3 - Deploy a PGD cluster

As with any other deployment in Kubernetes, to deploy a PGD cluster you need to apply a configuration file that defines your desired PGDGroup resources that make up a PGD cluster.

Some sample files are included in the EDB Postgres Distributed for Kubernetes repository. The flexible_3regions.yaml manifest contains the definition of a PGD cluster with two data groups and a global witness node spread across three regions. Each data group consists of two data nodes and a local witness node.

Regions and availability zones

When creating Kubernetes clusters in different regions or availability zones for cross-regional replication, ensure the clusters can communicate with each other by enabling network connectivity. Specifically, every service created with a -node or -group suffix must be discoverable by all other -node and -group services. You can achieve this by deploying a network connectivity application like Submariner on every cluster.

Further reading

For more details about the available options, see the "API Reference" section.

You can deploy the flexible-3-regions example by saving it first and running:

kubectl apply -f flexible_3regions.yaml

You can check that the pods are being created using the get pods command:

kubectl get pods

The pods are being created as part of PGD nodes. As described in Architecture, they're implemented on top of EDB Postgres for Kubernetes clusters.

You can list the clusters then, which shows the PGD nodes:

$ kubectl get clusters
NAME         AGE     INSTANCES   READY   STATUS                     PRIMARY
region-a-1   2m50s   1           1       Cluster in healthy state   region-a-1-1
region-a-2   118s    1           1       Cluster in healthy state   region-a-2-1
region-a-3   91s     1           1       Cluster in healthy state   region-a-3-1
…
…

Ultimately, the PGD nodes are created as part of the PGD groups that make up your PGD cluster.

$ kubectl get pgdgroups
NAME       DATA INSTANCES   WITNESS INSTANCES   PHASE                AGE
region-a   2                1                   PGDGroup - Healthy   4m50s
region-b   2                1                   PGDGroup - Healthy   4m50s
region-c   0                1                   PGDGroup - Healthy   4m50s

Notice how the region-c group is only a witness node.