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

# High-Level Deployment Steps

The deployment of a RegattaDB system follows the sequence below. This overview assumes that you have already completed the [**Prepare for Deployment**](/self-hosted-deployment/manual-deployment-copied-1/prepare-for-deployment) guide and provisioned your environment accordingly. The following sections summarize the deployment workflow. The next page provides a detailed, step-by-step deployment example covering the entire process.

During the deployment process, you will use SQL management commands to communicate with the **SM**. For a complete reference of the SQL management commands, see the [**System Management SQL Specification**.](/sql/system-monitoring/system-management/system-state)

### Identify the Deployment Nodes

Identify the nodes that will host the RegattaDB modules.

For each node, prepare:

* Internal IP addresses for cluster communication.
* External IP addresses for client application access (if required).

### Deploy the SM

RegattaDB uses a single **SM** (System Manager) module.

* Select the node that will host the **SM**.
* Install the **SM** package.
* Start the **SM** service.

### Deploy the SNA

Deploy an **SNA** on every node that will host one or more Operational Modules.

* Install the **SNA** package.
* Start the **SNA** service.

### Plan Module Placement

Determine which Operational Modules will run on each node. A single node may host one or more modules.

Each deployment requires:

* Exactly one **DCM**
* Exactly one **Sequencer**
* Exactly one **GDD**
* One or more **RDB** modules

### Install the Operational Modules

Install the RPM packages corresponding to the assigned module roles on each node.

### Enter SETUP State

Install the RegattaDB CLI, which is included in the software package. The CLI may be installed on any node that has network connectivity to the **SM**.

Transition the system from the **DOWN** state to the [**SETUP**](/sql/system-management/system-states-and-commands#system-enter-setup) state to enable configuration operations.

### Define the Infrastructure

Register the cluster nodes using their host names and configured IP addresses.

### Configure Storage

[Register the storage devices](/sql/system-management/storage-devices) assigned to each node by specifying a unique name and the local device path for each storage device.

### Add the Operational Modules

[Register each module](/sql/system-management/modules#module-add), assign it to its designated node, and associate it with the appropriate storage device.

<Note>
  By default, RegattaDB automatically allocates memory according to the requirements of each module. Because the **RDB** is designed to consume all remaining available memory, it should always be provisioned **after** the other modules. This ensures it receives the memory that remains after the fixed allocations for the other modules have been reserved. Although RegattaDB supports manual memory allocation on a per-module basis, the default allocation should be used unless your sizing analysis requires otherwise.
</Note>

### Exit SETUP State

Transition the system back to the [**DOWN**](/sql/system-management/system-states-and-commands#system-exit-setup) state to finalize and lock the deployment configuration.

### Activate the System

[Start the system](/sql/system-management/system-states-and-commands#system-start) to transition it to the **ACTIVE** state. During this step, the Operational Modules are started and begin accepting client connections.