How to Install Looker

Getting Looker up and running is the first step to unlocking its powerful data modeling and visualization capabilities. The installation process isn't a one-size-fits-all path, it has two distinct routes depending on how much control you want over the infrastructure. This guide will walk you through setting up a Looker instance, covering both the cloud-managed and self-hosted options so you can choose and execute the setup that's right for your team.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Choose Your Path: Google-hosted vs. Self-hosted

Before you get started, the most critical decision is where your Looker instance will live. This single choice determines the entire installation and maintenance workflow.

• Looker (Google Cloud Core): This is the fully managed, SaaS version of Looker hosted on Google Cloud Platform (GCP). Google handles all the underlying infrastructure, maintenance, and updates for you. This is the streamlined path, ideal for teams that want to get analyzing fast without worrying about server administration.
• Customer-Hosted Looker (Self-Hosted): With this option, you install and run the Looker application on your own infrastructure. This could be on-premise servers or your own virtual machines in a cloud like AWS, Azure, or GCP. This route offers maximum control and customizability but requires significant technical expertise to manage.

We'll break down the installation steps for both, starting with the simpler Google Cloud option.

Path 1: How to Install Looker (Google Cloud Core)

If you prefer a hands-off approach to server management, the Google Cloud-hosted solution is your best bet. The setup is handled almost entirely through the Google Cloud Console interface, turning a multi-day server setup into a process you can complete in under an hour.

Step 1: Check Your Prerequisites

Before creating an instance, you need a few things in place:

• A Google Cloud Project: All GCP resources live inside a project. You'll need an active project to house your Looker instance.
• The Right Permissions: Your Google Cloud user account needs the appropriate Identity and Access Management (IAM) role to create and manage Looker instances. The Looker Admin and Project Owner roles both have sufficient permissions.
• Enabled APIs: You may need to enable the Looker (Google Cloud Core) API and the Secret Manager API from the APIs & Services dashboard in your project.

Step 2: Create a New Looker Instance

With the prerequisites sorted, you can now provision your instance.

1. Navigate to the Google Cloud Console.
2. In the search bar at the top, type "Looker (Google Cloud Core)" and select it from the results.
3. Click the Create Instance button. This will take you to the main configuration page.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Step 3: Configure Your Instance Settings

This page asks for the core details about how your Looker instance will operate. Here's a quick rundown of each field:

• Instance Name: Give your instance a memorable name. This will also form part of its unique URL.
• Region: Select the geographic region where your instance will be hosted. For best performance, choose the region closest to your users or your data warehouse.
• Edition: Looker offers three editions:
• Network Connection: This is a critical security setting.

Once you are done with the configuration details, review your choices and click Create. Google will now begin provisioning your instance in the background. This can take anywhere from 30 minutes to an hour. Once complete, you'll see a green checkmark next to your instance name and an instance URL will be generated. You can click this URL to access your brand-new Looker environment and begin the final setup, like connecting to your database.

Path 2: The Customer-Hosted Installation

Choosing to host Looker on your own gives you complete dominion over the environment, from the operating system to the networking rules. This is the path for organizations that demand low-level control for compliance, data governance, or custom integrations. Be prepared - this process is far more technical.

Step 1: Gather Your Prerequisites

The self-hosted route has a much longer checklist of requirements.

• A Looker License Key: You cannot download or run the software without a license key provided by the Looker sales team.
• A Host Machine: You need a server (virtual or physical) that meets Looker's minimum system requirements. These change over time, so always check the official documentation, but generally expect to need at least:
• Java Development Kit (JDK): Looker runs on Java. You must install the specific version of Amazon Corretto or OpenJDK recommended in Looker's documentation.
• A Backend Database: Looker needs its own database to store application data - things like user information, saved reports (Looks), and dashboard configurations. MySQL is the standard recommendation, but Looker supports other databases like PostgreSQL. This is separate from the analytics database you'll be connecting to later.

Step 2: Prepare the Host Environment

Start by provisioning your server and installing the required software.

1. Install Java

After SSHing into your server, install the correct version of Java. For example, on Ubuntu, the command might look like this:

sudo apt-get update
sudo apt-get install openjdk-11-jdk-headless

2. Create a Looker User

For security, you should not run the Looker application as the root user. Create a dedicated system user and group for Looker:

sudo groupadd looker
sudo useradd -m -g looker looker

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Step 3: Set Up the Backend Database

Next, you must prepare the database that Looker will use for its own internal storage.

1. Log in to your MySQL server.
2. Create a database specifically for Looker. Be sure to use the utf8mb4 character set for full character support.

CREATE DATABASE looker_internal DEFAULT CHARACTER SET utf8mb4 DEFAULT COLLATE utf8mb4_general_ci,

1. Create a database user and grant it permissions to the new database. Remember to replace 'your_secret_password' with a strong, secure password.

CREATE USER 'looker_user'@'%' IDENTIFIED BY 'your_secret_password',
GRANT ALL PRIVILEGES ON looker_internal.* TO 'looker_user'@'%',
FLUSH PRIVILEGES,

Keep these credentials (database name, username, and password) handy. You will need them soon.

Step 4: Download and Deploy the Looker Application

Now you can place the actual Looker software on the server.

1. Your Looker account manager will provide you with a secure link to download the required Looker JAR files (looker.jar and looker-dependencies.jar).
2. Log in to the server as your new looker user (su - looker).
3. Create a directory for the Looker application, for example, ~/looker.
4. Download the JAR files from the secure link and place them in the ~/looker directory.

Step 5: Create Looker Configuration and Startup Scripts

This is the most critical technical step. To start, Looker needs to know where to find its backend database. Create a file named looker-db.yml in your ~/looker directory.

# ~/looker/looker-db.yml
host: your_database_host
port: 3306
database: looker_internal
username: looker_user
password: your_secret_password
dialect: mysql

Next, you'll create a script to launch the application. This script runs the Java command, passes it the database credentials, and allocates memory. Create a file named start-looker.sh in the ~/looker directory and add the following:

#!/bin/bash
# ~/looker/start-looker.sh

LOOKER_DIR=$(dirname "${BASH_SOURCE[0]}")

java -Xms4096m -Xmx4096m \
  -jar "${LOOKER_DIR}/looker.jar" \
  start \
  --mysql-use-legacy-utf8mb4-config \
  --no-ssl

A few notes on this script:

• -Xms4096m -Xmx4096m sets the minimum and maximum heap size (memory) for the Java Virtual Machine. Adjust this based on your server's available RAM.
• --mysql-use-legacy-utf8mb4-config instructs Looker to find its database credentials in the looker-db.yml file from the previous step. There are other ways to pass credentials, but this is a common method.

Finally, make the script executable:

chmod +x start-looker.sh

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Step 6: Launch Looker

You're ready to start the application. Simply run your startup script:

./start-looker.sh

The first time you run this, Looker will perform its internal database migrations, which may take several minutes. Once you see a successful launch message in your terminal, Looker is running! You can now access it in a browser by navigating to http://your-server-ip:9999.

Final Thoughts

Installing Looker means making a fundamental choice between convenience and control. Using the managed Google Cloud service allows you to be operational within an hour, while the customer-hosted route provides ultimate authority over your BI environment at the cost of significant setup and ongoing maintenance.

For organizations without a dedicated infrastructure team, this level of setup can feel overwhelming. After all, the goal is to get answers from your data, not become server administrators. We built Graphed for teams who want to bypass this friction entirely. It connects to your data sources with a few clicks and lets you build real-time dashboards using simple, conversational language, freeing you up to focus on insights instead of installations.