Database configuration for Neptune DXP - Open Edition
All data, ranging from your configuration, apps, and application data, is kept in the Neptune DXP - Open Edition database.
Neptune DXP - Open Edition supports the following databases:
-
Microsoft SQL Server
-
PostgreSQL
-
SQLite
By default, the embedded SQLite database is configured and used. SQLite is suited to evaluation and development. For a production setup, use Microsoft SQL Server or PostgreSQL.
| Neptune does not recommend SQLite for production. |
|
The data from your previously installed Neptune DXP - Open Edition database is deleted during the installation process of the new Neptune DXP - Open Edition version. Perform the following steps to avoid deletion of your data:
|
Microsoft SQL Server
The AI Context Hub and
vectorization need the native
VECTOR data type, which requires SQL Server 2025 or later. Earlier versions have
no vector support.
|
PostgreSQL
-
Create a new database.
-
Create the planet9 schema.
CREATE DATABASE planet9;
CREATE SCHEMA planet9;
On initialization, Neptune DXP - Open Edition creates all the needed tables under this schema.
Vector support with pgvector
The AI Context Hub and vectorization need the pgvector extension. For the AI Context Hub, install and enable it before the first start of a Neptune DXP - Open Edition version that includes the AI Context Hub, version 25.0 or later.
-
Install pgvector on the PostgreSQL server. How you install it depends on how you run PostgreSQL. For example, the pgvector Docker image already includes the extension. For all options, see the pgvector installation documentation.
-
In the
planet9database, create theplanet9schema if it does not exist yet, and enable the extension:CREATE SCHEMA IF NOT EXISTS planet9; CREATE EXTENSION IF NOT EXISTS vector;
If you enable pgvector after the AI Context Hub tables already exist, re-create the embedding columns. See Enabling vector support for the AI Context Hub.