#Create an Instance

#Feature Overview

This feature allows users to create and manage PostgreSQL database instances within a Kubernetes cluster. Users can quickly deploy database instances that meet their needs by configuring:

• Resource specifications
• Account information
• Storage requirements

#Prerequisites

Before creating a PostgreSQL instance, please ensure that:

1. Appropriate storage classes have been configured.
2. Postgres Operator is correctly installed and running.
3. You have the necessary permissions to create resources.

#Version Support

Currently, PostgreSQL versions 11,12,14 are supported, with the recommendation to use the latest stable version.

#Architecture Choices

Postgres Operator supports the following architectures:

• Single Node: Suitable for development and testing environments
• Primary-Replica Replication: Recommended for production environments with:
  • Automatic failover
  • High availability
  • Data redundancy

#Procedure

# Example minimal HA cluster configuration
cat << EOF | kubectl create -f -
apiVersion: acid.zalan.do/v1
kind: postgresql  # Custom resource definition for PostgreSQL clusters
metadata:
  name: pg-single
  namespace: c1-midautons
spec:
  ipFamilyPrefer: ""
  teamId: ACID
  enableExporter: true
  enablePgpool2: false
  spiloPrivileged: false
  spiloRunAsGroup: 103
  spiloRunAsUser: 101
  spiloAllowPrivilegeEscalation: false
  enableReadinessProbe: true
  postgresql:
    parameters:
      log_directory: /var/log/pg_log
    version: "14"
  numberOfInstances: 1
  resources:
    requests:
      cpu: "1"
      memory: 2Gi
    limits:
      cpu: "1"
      memory: 2Gi
  volume:
    size: 5Gi
    storageClass: c1-topolvmsc
EOF

cat << EOF | kubectl -n $NAMESPACE create -f -
apiVersion: acid.zalan.do/v1
kind: postgresql
metadata:
  name: pg-ha
  namespace: default
spec:
  ipFamilyPrefer: ""
  teamId: ACID
  enableExporter: true
  enablePgpool2: false
  spiloPrivileged: false
  spiloRunAsGroup: 103
  spiloRunAsUser: 101
  spiloAllowPrivilegeEscalation: false
  enableReadinessProbe: true
  postgresql:
    parameters:
      log_directory: /var/log/pg_log
    version: "14"
  numberOfInstances: 3
  resources:
    requests:
      cpu: "1"
      memory: 2Gi
    limits:
      cpu: "1"
      memory: 2Gi
  volume:
    size: 50Gi
    storageClass: default
  patroni:
    ttl: 30
    loop_wait: 10
    retry_timeout: 10
EOF

kubectl -n $NAMESPACE get postgresql

NAME                            AGE
pg-ha                           21h

Note

1. For production environments, use the primary-replica architecture
2. Verify your storage class supports dynamic provisioning before deployment
3. Configure appropriate resource limits
4. Implement regular backup procedures for critical data

Warning

• Improper resource limits may lead to performance issues or pod evictions
• Without backups, data loss may occur during failures