Hardware and Software Requirements

This page details the system requirements and recommendations for running TigerGraph in a production or development environment.

Hardware Requirements

  • CPU: The TigerGraph database runs on x86_64 CPUs, with a minimum of 4 cores. See the Hardware Recommendations section for additional considerations.

  • Disk Storage: Solid State Disk (SSD) storage is required, not just to improve performance, but also to avoid timeout and other timing issues.

Certified Operating Systems

The TigerGraph Software Suite is designed to run on 64-bit Linux. It can run on a variety of Linux 64-bit distributions. The software has been tested on the operating systems listed below:

Table 1. List of certified OS
Distro Supported

RedHat 7.0 to 8.3

CentOs 6.5 to 8.0

Amazon Linux 2016.03 to 2018.03

Ubuntu 16.04 LTS

Ubuntu 18.04 LTS

Ubuntu 20.04 LTS

Debian 8

Suse 12

Oracle Linux 8.0 to 8.4

Windows (all version)


MacOS (Intel and M1 chip)


When a range of versions is given, it means that the software has been tested on the oldest and newest versions. We continually evaluate the operating systems on the market and work to update our set of supported operating systems as needed.

The TigerGraph installer installs its own copies of Java JDK and GCC, accessible only to the TigerGraph Linux user, to avoid interfering with any other applications on the same server.

Prerequisite Software


Use a bash shell for the installation process.


Before offline installation, the TigerGraph system needs a few basic software packages to be present:

  • crontab

  • curl

  • ip

  • more

  • netcat

  • netstat

  • net-tools

  • ssh/sshd (Only required for cluster installation)

  • sshpass

  • tar

If they are not present, the TigerGraph installer will exit with an error reporting which packages are missing. Contact your system administrator to have them installed on your target system. For example, they can be installed with one of the following commands.
# Centos or RedHat:
sudo yum install tar curl cronie iproute util-linux-ng net-tools nc coreutils openssh-clients openssh-server sshpass

# Ubuntu or Debian (Except Ver 18.04):
sudo apt install tar curl cron iproute util-linux net-tools netcat coreutils openssh-client openssh-server sshpass

# Ubuntu or Debian (Ver. 18.04):
sudo apt install tar curl cron iproute2 util-linux net-tools netcat coreutils openssh-client openssh-server sshpass


If you are running TigerGraph on a multi-node cluster, you must install, configure and run the NTP (Network Time Protocol) daemon service. This service will synchronize system time among all cluster nodes.


If you are running TigerGraph on a multi-node cluster, you must configure the iptables/firewall rules to make all TCP ports open among all cluster nodes.


In an on-premises installation, the system is fully functional without a web browser. To run the optional browser-based TigerGraph GraphStudio User Interface or Admin Portal, you need an appropriate browser:

Browser Chrome Safari Firefox Opera Edge Internet Explorer

Supported version







Other browser-based products, such as TigerGraph Insights, have their own browser requirements. See their corresponding documentation.

Hardware Recommendations

Choosing the right hardware to host your TigerGraph system is crucial for the right balance of cost and performance. This page provides some general guidelines for hardware selection based on simple hypothetical assumptions, but your actual hardware requirements will vary based on your data size, workload, and performance requirements.

The sizing recommendations below apply to each server node. If you have more than several hundred gigabytes of data, you should consider deploying a cluster of multiple nodes, to distribute your data.

Consult a TigerGraph solution architects for an estimate of memory and storage needs.

Personal Use

Example profile: A single user on a personal computer (desktop or laptop), who is learning to use the software and to write applications.

Component Specification


4 cores (≥ 2.10GHz)


8 GB


≥ 50 GB (SSD)


1 Gigabit Ethernet adapter

Development, UAT, or SIT System

Component Specification


≥ 16 cores (≥ 3.00GHz)


≥ 32 GB


≥ 300 GB (SSD)


1 Gigabit Ethernet adapter

Production System

Production systems should consider having replica clusters for redundancy and availability.
Component Specification


≥ 32 cores (≥ 3.40GHz)


≥ 64 GB


≥ 500 GB (SSD)


10 Gigabit Ethernet adapter

Additional Considerations

This section describes considerations beyond sizing for the CPU, memory, and storage.


Choosing the right CPU model is important. For example, Intel Core i5 may be good enough for development or certain test systems where latency and throughput are not critical. For production systems, we suggest higher grade CPUs such as Intel Core i7, Intel Xeon Gold, and AMD Epyc.

CPU clock speeds are not directly comparable across different processor designs, but within the same processor family, faster is better. We highly recommend using a speed of 3.00GHz or higher.

Below is an example of a production hardware configuation.

Architecture:        x86_64
CPU op-mode(s):      32-bit, 64-bit
Byte Order:          Little Endian
CPU(s):              24
On-line CPU(s) list: 0-23
Thread(s) per core:  1
Core(s) per socket:  1
Socket(s):           24
NUMA node(s):        1
Vendor ID:           GenuineIntel
CPU family:          6
Model:               85
Model name:          Intel(R) Xeon(R) Gold 6248R CPU @ 3.00GHz
Stepping:            7
CPU MHz:             2992.969

RAM (Memory)

RAM is needed both to store a working copy of our data and to do computational work. TigerGraph is a hybrid in-memory database, meaning that in order to achieve maximum performance, it will attempt to load a copy of your entire working graph into RAM. The RAM size is based on multiple factors:

  • Type of workload (e.g., OLAP/OLTP queries, intensive CRUD operation…​.)

  • Query SLAs

  • Dataset partitioning (number of partitions in your TigerGraph cluster)

RAM is not only size but also type. DDR4 or faster is recommended because it gives you significantly faster data transfer rates, better power efficiency, and better data granularity than DDR3.


As stated in the Hardware Requirements section, solid state drive (SSD) storage is mandatory for TigerGraph. The longer data access times of spinning hard disks (HDD) can cause several issues, such as the following:

  • Timeouts between internal components of TigerGraph

  • Issues when making a backup of a TigerGraph data

  • Issues when restoring a TigerGraph database

  • Low performance

We recommend using different mount points for the App, Data, Log and Temp directories to avoid one single point of failure. This can be defined while installing TigerGraph (either via interactive mode or non-interactive mode). Here is an example of how to specify different mount points in the install_config.json file.

"RootDir": {
"AppRoot": "/mnt1/app",
"DataRoot": "/mnt2/data",
"LogRoot": "/mnt3/log"
"TempRoot": "/mnt4/tmp"
To economize, you can opt to use a magnetic hard disk (HDD) for the logs; the remaining ones must be SSD.

Another important point when choosing the Disk type is the disk IOPS capacity. We strongly suggest opting for ≥3000 IOPS.