> ## Documentation Index
> Fetch the complete documentation index at: https://docs.qodo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Set up a database

> Qodo Context Engine requires a PostgreSQL database to store its indexing data and metadata. This setup enables fast, reliable retrieval of code context and insights across your repositories.

## Recommended instance configuration

For most organizations, we recommend the following Postgres instance sizing:

* **4 vCPU**
* **32 GB RAM**
* **100 GB Disk**

This setup works for most use cases. You can experiment with smaller sizes if your codebase is relatively small.

## Database requirements

* **PostgreSQL 17** (or later)
* `pg_vector` **extension installed**

## Setup instructions

**1. Allow network access**

* Enable **TCP connections on port 5432**
* Whitelist access **from the Kubernetes cluster** where Qodo Context Engine will be deployed

**2. Create two databases**

* One for **RAG Indexer**
* One for **Metadata**

**3. Create two users**

* Each user should have **full access (read/write)** to its respective database

**4. Whitelist Qodo’s server IP**

* Qodo will provide the IP address of our **STS server**
* Please **whitelist it** to allow secure access

## Share credentials with Qodo

Send the following details securely to your Qodo contact:

**RAG Indexer DB:**

```bash theme={null}
host = ""
port = ""
username = ""
password = ""
db_name = ""
```

**Metadata DB:**

```bash theme={null}
DATABASE_USER = ""
DATABASE_PASSWORD = ""
DATABASE_NAME = ""
DATABASE_HOST = ""
DATABASE_PORT = ""
```


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.