Installation on Bare Metal

This guide describes how to install TigerGraph on bare-metal infrastructure.

The installation guide covers installation as a single node or as a multi-node cluster, interactively or non-interactively.

If you signed up for the Enterprise Free license, you also have access to the TigerGraph platform as a Docker image or a virtual machine (VirtualBox) image. Follow the instructions in Getting started to start up TigerGraph in a Docker container or with VirtualBox.

1. Prerequisites

Your machines meet the minimum Hardware and Software Requirements.
- The installation machine can be within or outside the cluster. If outside the cluster, the installation machine still needs to be a Linux machine.
A license key provided by TigerGraph, unless you are using Enterprise Free.
A TigerGraph system package.
Every machine in the cluster has a Linux superuser with the same username and password, or same username and SSH key.
If you are installing a cluster, ensure that every machine has the same SSH port and that the port stays open during installation.
- SSH port does not need to be 22. You can specify a custom port number during installation.
All ports used for internal communication must be allowed for a correct TigerGraph installation and usage.
All provisioned machines must be in the same region if installing on cloud.
All provisioned machines must have the same CPU, RAM and disk size.
All provisioned machines must have enough disk space available (≥ 50 GB) for TigerGraph Cluster installation.
All provisioned machines must be running SSD disk types.
All provisioned machines must have all clocks in-sync.
All provisioned machines must allow password-less SSH for the Linux superuser.
All provisioned machines must be serving only TigerGraph.

1.1. For Oracle Linux and RedHat distros

If you install on Oracle Linux Server (OEL) or RedHat (RHEL), there are a few additional prerequisites.

For each node, create an OS user and password.
```
sudo useradd NEW-USER
sudo passwd NEW-USER
```
Add the new user to the AllowUsers row in the /etc/ssh/sshd_config file.
```
AllowUsers OLD-USER NEW-USER
```
Restart sshd.
```
/sbin/service sshd restart
```

This adds firewall rules to allow TCP communication between nodes in the cluster. The installation process will prompt you for these steps. If you have already completed them, ignore and continue.

2. Non-interactive installation

In non-interactive installation, you specify all the desired configurations in a JSON file, and run the installation script with the configurations specified in the file. This is the recommended approach as it allows you to see all your configurations in one place and diagnose issues more easily.

2.1. Extract package

Extract the package by running the following command to create a folder named tigergraph-<version>-offline. The filename of your package may vary depending on the product edition and version.

$ tar -xvzf tigergraph-<version>.tar.gz

2.2. Edit configuration file

Navigate to the tigergraph-<version>-offline folder and open the configuration file called install_conf.json.

Edit the configuration file according to your needs, pay extra attention to the followings entries:

TigerGraph Username and Password: A TigerGraph database has a built-in superuser, whose default username and password are tigergraph. You can either customize the username and password now or after installation. If the username already exists, the installer will skip creating a new user.

Deploying a database without changing the password for the TigerGraph superuser is a security risk.

SSHPort: By default, the SSH port is 22, if this port is open in all the provisioned machines you can leave it as it is. Otherwise change the port to the actual SSH port in your environment. The SSH port number must be the same across all the provisioned machines.
PrivateKeyFile and PublicKeyFile: The TigerGraph installer creates these keys by default, so these fields are optional. If you want to pass your own keys you can add the absolute paths here.
NodeList: The list of your provisioned machines IPs in the following JSON format:
```
    "NodeList": [
      "m1: 123.456.78.99",
      "m2: 123.456.78.98",
      "m3: 123.456.78.97",
      "m4: 123.456.78.96"
    ]
```
- For single server installation, you only specify the IP address for m1. You can either use loop-back address (e.g. 127.0.0.1) or the public IP of the machine. if you use loop-back address IP for m1, you do not need to specify any sudo user for SudoUser or authentication method in the Method section.
SudoUser: Username of the sudo user who will be used to execute the installation on all nodes. This user must exist on all the provisioned machines
Authentication Method: This refers to how the above sudo user will authenticate between the provisioned machines. It can authenticate via password (chose the P method and pass the password) or via SSH key (choose the K method and pass the absolute path for the SSH key).
ReplicationFactor: The number of replicas in your cluster. Refer to Replication factor for detailed description.

2.2.1. Example

The following is an example configuration:

{
  "BasicConfig": {
    "TigerGraph": {
      "Username": "mark",
      "Password": "markpassword",
      "SSHPort": 22,
      "PrivateKeyFile": "",
      "PublicKeyFile": ""
    },
    "RootDir": {
      "AppRoot": "/home/tigergraph/tigergraph/app",
      "DataRoot": "/home/tigergraph/tigergraph/data",
      "LogRoot": "/home/tigergraph/tigergraph/log",
      "TempRoot": "/home/tigergraph/tigergraph/tmp"
    },
    "License": "<license>",
    "NodeList": [
      "m1: 123.456.78.99",
      "m2: 123.456.78.98",
      "m3: 123.456.78.97",
      "m4: 123.456.78.96"
    ]
  },
  "AdvancedConfig": {
    "ClusterConfig": {
      "LoginConfig": {
        "SudoUser": "tom",
        "Method": "K",
        "P": "<sudo_user_password>",
        "K": "/home/tom/mykey.pem"
      },
      "ReplicationFactor": 2
    }
  }
}

2.3. Run installation script

Save the configuration changes and run sudo ./install.sh -n to install the TigerGraph Cluster in non-interactive mode based on the configuration you passed in the install_conf.json file.

3. Interactive installation

With interactive installation, you do not need to edit the install_conf.json configuration file as you will be asked interactively to pass the required information while you install TigerGraph.

3.1. Extract package

Extract the package by running the following command to create a folder named tigergraph-<version>-offline. The filename of your package may vary depending on the product edition and version.

$ tar -xvzf tigergraph-<version>.tar.gz

3.2. Run installation script

Navigate to the tigergraph-<version>-offline` folder and run the install.sh script with the following commands:

$ cd tigergraph-<version>-offline

#sudo not requried for standalone installation
$ sudo ./install.sh

Provide the information that the installer asks for. You may choose to hit Enter to skip and use the system default or enter a new value:

Your agreement to the License Terms and Conditions
Your license key (not applicable for Enterprise Free)
Username for the Linux user who will own and manage the TigerGraph platform
- The installer creates a Linux user with this username who is the only authorized user that can run gadmin commands to manage the TigerGraph Platform.
- If the installation and data folders are modified from their defaults, the newly created user must have appropriate permissions for these folders. The installation folder requires read/write/execute access, and the other folders require read/write access.
Password for the Linux user who will own and manage the TigerGraph platform
Path to where the installation folder will be
Path to where the data folder will be
Path to where the log folder will be
- For production systems, it’s recommended that you use a separate disk partition to store log files to prevent out-of-space issues.
Path to where the temp folder will be
The SSH port for your machine