New to KubeDB? Please start here.

Configuring Highly Available PostgreSQL Cluster

In PostgreSQL, multiple servers can work together to serve high availability and load balancing. These servers will be either in Master or Standby mode.

In master mode, server that can modify data. In standby mode, the server continuously applies WAL received from the master server. The standby server can read WAL from a WAL archive (see restore_command) or directly from the master over a TCP connection (streaming replication).

Standby servers can be either warm standby or hot standby server.

Warm Standby

A standby server that cannot be connected to until it is promoted to a master server is called a warm standby server. Standby servers are by default warm standby unless we make them hot standby.

The following is an example of a Postgres object which creates PostgreSQL cluster of three servers.

apiVersion: kubedb.com/v1alpha1
kind: Postgres
metadata:
  name: warm-postgres
  namespace: demo
spec:
  version: "9.6-v2"
  replicas: 3
  standbyMode: Warm
  storageType: Ephemeral

In this examples:

• This Postgres object creates three PostgreSQL servers, indicated by the replicas field.
• One server will be primary and two others will be warm standby servers, as instructed by spec.standbyMode

Hot Standby

A standby server that can accept connections and serves read-only queries is called a hot standby server.

The following Postgres object will create PostgreSQL cluster with hot standby servers.

apiVersion: kubedb.com/v1alpha1
kind: Postgres
metadata:
  name: hot-postgres
  namespace: demo
spec:
  version: "9.6-v2"
  replicas: 3
  standbyMode: Hot
  storageType: Ephemeral

In this examples:

• This Postgres object creates three PostgreSQL servers, indicated by the replicas field.
• One server will be primary and two others will be hot standby servers, as instructed by spec.standbyMode

High Availability

Database servers can work together to allow a second server to take over quickly if the primary server fails. This is called high availability. When primary server is unavailable, standby servers go through a leader election process to take control as primary server. PostgreSQL database with high availability feature can either have warm standby or hot standby servers.

To enable high availability, you need to create PostgreSQL with multiple server. Set spec.replicas to more than one in Postgres.

Load Balancing

Master server along with standby server(s) can serve the same data. This is called load balancing. In our setup, we only support read-only standby server. To enable load balancing, you need to setup hot standby PostgreSQL cluster.

Read about hot standby and its setup in Postgres.

Replication

There are many approaches available to scale PostgreSQL beyond running on a single server.

Now KubeDB supports only following one:

• Streaming Replication provides asynchronous replication to one or more standby servers.

These standby servers can also be hot standby server. This is the fastest type of replication available as WAL data is sent immediately rather than waiting for a whole segment to be produced and shipped.

KubeDB PostgreSQL support [Streaming Replication](/docs/v0.13.0-rc.0/guides/postgres/clustering/streaming_replication)

Next Steps

• Learn how to setup Streaming Replication