Self-Hosted Remote Deployment of the Zoho Analytics MCP Server

The self-hosted remote deployment option for the Zoho Analytics MCP Server allows your organisation to run a single, centrally managed server instance that all users share via HTTPS. Unlike local setups where each user runs their own process, this approach centralises authentication, updates, and credential management on a single host machine — reducing overhead and simplifying governance across your team.

How This Differs from Local Installations

In Docker and NPM local setups, every user runs and maintains their own MCP server process, along with their own credentials. The remote self-hosted approach deploys one shared instance on a host machine that serves all users from a central point. This means credential management, version updates, and access controls are handled once at the server level rather than repeated across individual machines.

Step 1: Register an OAuth Client

Before deploying the server, you must register an OAuth application in the Zoho Developer Console. This provides the credentials needed for the server to authenticate users on your behalf.

Open the Zoho Developer Console. Navigate to the console and create a new application.
Select Server-based Application. This application type supports redirect-based OAuth flows, which the remote MCP server requires.
Set the redirect URI. Enter <your-public-url>/auth/callback as the redirect URI, replacing the placeholder with your actual public HTTPS URL.
Save your credentials. Note down the Client ID and Client Secret — you will need both during configuration.

Note: Do not use a Self-Client application type for the remote MCP setup. Self-Client applications do not support redirect-based OAuth flows and will not work with this deployment method.

Step 2: Configure Environment Variables

The remote MCP server is configured through environment variables. Create a .env file containing the following settings.

Required Variables

Variable	Description
`ANALYTICS_SERVER_URL`	The base URL for the Zoho Analytics API
`OIDC_PROVIDER_CLIENT_ID`	The OAuth Client ID from the Developer Console
`OIDC_PROVIDER_CLIENT_SECRET`	The OAuth Client Secret from the Developer Console
`MCP_SERVER_PUBLIC_URL`	The public HTTPS URL where your server is accessible
`SESSION_SECRET_KEY`	A random string used to encrypt session data
`MCP_SERVER_ORG_IDS`	Comma-separated list of Zoho Analytics organisation IDs permitted to access this server

Optional Variables

You may also configure the PORT the server listens on, STORAGE_BACKEND (defaulting to in-memory), and Redis connection details if using Redis as a session store.

Production recommendation: Use Redis as your storage backend rather than the default in-memory store. In-memory storage does not persist across server restarts and is unsuitable for production environments.

Step 3: Deploy the Container

With your .env file in place, start the server using the official Docker image:

docker run --name remote-mcp-server \
  --network=host \
  --env-file .env \
  zohoanalytics/mcp-server:remote-v1

The --network=host flag allows the container to use the host machine's network stack, which is required for the server to be reachable at its configured public URL.

Step 4: Configure Client Connections

Once the server is running, each user connects to it by adding an entry to their AI client's MCP host configuration file. The exact file location varies by application, but the entry follows this format:

{
  "mcpServers": {
    "ZohoAnalyticsMCP": {
      "url": "<hosted-mcp-server-url>/mcp",
      "type": "http"
    }
  }
}

Replace <hosted-mcp-server-url> with the public HTTPS URL you configured in your environment variables. Users do not need to install anything locally — the AI client connects directly to your hosted server.

Available MCP Tools

The deployed server exposes 21 tools to connected AI clients, covering workspace and table creation, data import and export, SQL query execution, report generation, and row-level operations against your Zoho Analytics data.

Support

Zoho Analytics provides 24x5 technical support by email at support@zohoanalytics.com, with toll-free telephone support available for users in the US, UK, Australia, and India.

Need help? 1 Cloud Consultants can assist with planning, configuring, and securing your self-hosted Zoho Analytics MCP Server deployment. Book a discovery call with 1 Cloud Consultants.

Zoho Analytics MCP Server Self-Hosted Remote Deployment Guide