Core concepts

The core concepts of the PostgreSQL service are described below.

Concept	Description
Instance group	Object that manages the lifecycle of instances and backups.
Instance	A virtual machine (VM) managed by an instance group. Role is determined as Primary or Standby based on availability.
Instance type	Virtualized hardware resource including vCPU and memory. - Various types are available to optimize for different workloads.
Instance/Instance group lifecycle	Enables users to check and manage the operational and task states based on the lifecycle stages.
Backup	Managed online without downtime. Includes creation, deletion, and usage of backups.
Volume	The database uses block storage volumes. - Users do not need to manually create or configure block storage.
Network & Security	Provides logically separated service environments using VPC network and security features. - All instances are provisioned in a VPC.
IAM & Security	IAM is used to assign appropriate access roles and securely manage PostgreSQL resources.

Instance group

An instance group is an object that manages the lifecycle of PostgreSQL instances and backups.
It has various status values, including operational and terminal states, which users can use to monitor and manage the group.

Instance group lifecycle and status values

When you create a PostgreSQL instance group, it starts in the Pending state, proceeds to Provisioning and Primary Available, and finally transitions to the Available state. At this point, users can access the instance group via the configured endpoint and port. The default ports are 5432 for Primary and 5433 for Standby, but users can specify ports between 1024 and 65535.

Lifecycle diagram

PostgreSQL instance group lifecycle

Status	Description
Pending	Preparing to create the instance group. - This is the initial state after creation.
Provisioning	Creating instances and configuring required packages.
Primary Available	At least one Standby instance is unavailable, but the Primary is operational. - HA setup remains intact.
Switching	During failover, a new Standby is promoted to Primary. - Triggered by manual or automatic failover.
Available	All instances are running and the group is accessible via endpoint. - Default ports: 5432 (Primary), 5433 (Standby).
Terminating	Deleting all instances and related resources in the group.
Terminated	All group resources are deleted.
Error	Temporary failure in some instances.
Fatal	Irrecoverable error during setup, such as VM creation failure.

Instance

An instance is a VM managed by an instance group.
PostgreSQL instances are virtualized environments including CPU and memory. Each instance has a defined role and status.

Instance roles

Role	Description
Primary	Handles read/write operations. - Accessed through the group endpoint.
Standby	HA standby instance that also serves as a read replica. - Created only when HA is selected.

Instance types

Select instance types based on your workload requirements. Each type offers different compute resources, and billing is based on usage.

Supported types: m2a (general-purpose), r2a (memory-optimized)

Instance Type	vCPU	Memory (GiB)
m2a.large	2	8
m2a.xlarge	4	16
m2a.2xlarge	8	32
m2a.4xlarge	16	64
m2a.8xlarge	32	128
m2a.12xlarge	48	192
m2a.16xlarge	64	256
m2a.24xlarge	96	384
r2a.large	2	16
r2a.xlarge	4	32
r2a.2xlarge	8	64
r2a.4xlarge	16	128
r2a.8xlarge	32	256
r2a.12xlarge	48	384
r2a.16xlarge	64	512
r2a.24xlarge	96	768

Instance lifecycle and status values

Instances start in Pending, proceed through Provisioning and Starting, and finally reach Running.

PostgreSQL instance lifecycle

Status	Description
Pending	Preparing to create the instance.
Starting	Allocating resources and starting the instance.
Restarting	Restart in progress, usually triggered by user.
Warning	Operational but with some limitations.
Running	Fully operational.
Volume-Extending	Volume expansion in progress.
Terminating	Instance is being deleted.
Terminated	Instance and resources are deleted.
Error	Service is disrupted due to instance-level failure.
Fatal	Irrecoverable failure during creation.

Availability and roles

Availability Mode	Instance Count	Description
Single	1	Creates only a Primary instance.
High Availability (HA)	Up to 6	Creates 1 Primary and 1–5 Standby instances.

Availability status

Availability status indicates replication health in HA mode.

PostgreSQL instance availability

Status	Description
Configuring	Replication setup in progress.
Configured	Replication is functioning normally.
Unavailable	Replication has stopped.

Backup

Backups are managed at the instance group level.

Backup lifecycle and status values

Backups go from Pending → Processing → Available.

PostgreSQL backup lifecycle

Status	Description
Pending	Backup request submitted.
Processing	Backup is in progress.
Available	Backup is complete and usable.
Restoring	Being used for restore. Cannot delete.
Deleting	Backup is being deleted.
Deleted	Backup is fully removed.
Error	Backup failed and is not usable.
Fatal	Backup creation failed completely.

Failover

PostgreSQL supports high availability with failover. If a failure occurs, failover promotes a Standby instance to Primary. Both automatic and manual failover are supported.

caution

Failover may cause a short service interruption.

Failover status

Status	Description
SWITCHING	Failover in progress. Service is not available.
PRIMARY-AVAILABLE	Standby configuration is in progress.

Volume

Uses block storage volumes for PostgreSQL data. No user configuration required.

Network and security

All PostgreSQL instances operate in a VPC. Use VPC features to isolate and secure environments based on your network requirements.