Installing OmniSci on Local Hardware
The links below take you to a page with an end-to-end installation script for a specific operating system and version of OmniSci.
The CPU (no GPUs) install does not support backend rendering. Pointmap and Scatterplot charts are not available. The GPU install supports all chart types.
- Requirements
- Docker Install
- CentOS/RHEL Install with Yum
- Ubuntu/Debian Install with Apt
- CentOS/RHEL Install with Tarball
- Ubuntu Install with Tarball
Requirements
- Operating Systems
- CentOS/RHEL 7.0 or later
- Ubuntu 16.04 or later
- Additional Components
- Headless JDK
- EPEL
wgetorcurl- Kernel headers
- Kernel development packages
- NVIDIA hardware and software (for GPU installs only)
- Hardware: Volta, Pascal, or Kepler series GPU cards. OmniSci recommends that each GPU card in a server or distributed environment be of the same series.
- Software: NVIDIA CUDA Drivers, version 384.81 or later. Run
nvidia-smito determine the currently running driver.
Docker Install
The Docker install requires the least preparation and configuration of the available installation choices.
CentOS/RHEL Install with Tarball
Tarball installation is available when other options are not accessible.
Community Edition
To download a specific tarball for your environment, see Download OmniSci Community Edition.
Ubuntu/Debian Install with Tarball
Tarball installation is available when other options are not accessible.
Community Edition
To download a specific tarball for your environment, visit Download OmniSci Community Edition.