Support for Operating Systems and Environments in SAS® Viya® 3.5
Operating System Support
SAS Viya software can be deployed on the operating systems that are listed on this page. Deploy the software on the operating system that you selected when you placed your software order.
End users can also access the product user interfaces for SAS Viya applications from a desktop computer, using a supported web browser. Because SAS software is not installed on this machine, the requirements are minimal. UNIX 64-bit and Windows 64-bit operating systems are supported. Apple OS X 10.10 and later is supported for SAS Viya user interfaces that support Apple Safari.
For information about web browsers and mobile platforms that are supported for client access, see the SAS Viya Support for Web Browsers page.
SAS Support for Alternative Operating Systems
See http://support.sas.com/techsup/pcn/altopsys.html for the statement of SAS’s support for operating systems other than those listed here.
Supported Distributions of Linux (for 64-bit x86)
In a multi-machine deployment, SAS recommends that all server machines have the same version of Linux, including the distribution, release, and patch level. A mixture of Linux versions is not supported. In addition, all CAS Server machines must be running the same version of Linux.
The following operating systems are supported:
- Red Hat Enterprise Linux 6.7 (64-bit) and later within 6.x
- Red Hat Enterprise Linux 7.1 and later within 7.x
- Oracle Linux 6.7 and later within 6.x
- Oracle Linux 7.1 and later within 7.x
- SUSE Linux Enterprise Server 12.2 and later within 12.x *
* SUSE Linux is not supported for SAS Business Orchestration Services, SAS Visual Investigator, or SAS Intelligence and Investigation Management deployments. SUSE Linux 15 is not supported.
Support for Linux on Power
Linux on the Power Architecture requires Red Hat Enterprise Linux 7.6 with the kernel version 4.14. The IBM POWER9 CPU is required.
SAS Viya 3.5 has been compiled specifically for the POWER9 architecture. The required kernel version 4.14 is provided by the Linux kernel-alt packages.
SAS Viya GPU support extends to Linux on Power.
The following SAS Viya products cannot be installed on Linux on Power:
SAS/ACCESS to Google BigQuery, SAS Analytics for IoT, SAS Assortment Planning, SAS Business Orchestration Services, SAS Demand Planning, SAS Event Stream Processing, SAS Financial Planning, SAS Intelligence and Investigation Management, SAS Markdown Optimization, and SAS Visual Investigator.
Supported Distributions of Windows
For SAS Viya and SAS Event Stream Processing:
- Windows Server 2012 R2 Datacenter, Essentials, Foundation, and Standard Edition (64-bit version)
- Windows Server 2016 Datacenter, Essentials, and Standard Edition (64-bit version)
- Windows Server 2019 Datacenter and Standard Edition (64-bit version)
SAS Viya requires Ansible version 2.7.2 through 2.9.x.
The Ansible controller machine must be connected to the Red Hat Network. Oracle Linux machines require an Oracle Linux support subscription.
SAS recommends using pip to install one of the supported versions of Ansible. However, if you want to install Ansible from Red Hat channels, the Ansible controller requires a Server-Optional subscription in addition to the Base (operating-system) subscription. Be aware that these channels do not always contain a supported version of Ansible. The managed nodes must also be registered to the Red Hat Network, but a Base subscription is sufficient.
With Oracle Linux 6 Update 5 or later, the ULN registration procedure automatically registers each host with the latest channels for the base repository and the Unbreakable Enterprise Kernel Release 3 (UEK R3).
The Ansible controller requires Python 2.6 or later. However, Python 2.x has reached end-of-life as of 01 January 2020. SAS recommends that you upgrade to Python 3.5 or later. Any machines in your deployment that you plan to manage with Ansible 2.5 or later also require Python 3.5 or a later release. The Ansible controller also requires the following Python modules, which are provided by the Extra Package for Enterprise Linux (EPEL) if the machine is registered:
- Jinja2 2.7
SAS Deep Learning Python API (SAS DLPy) requires Python 3.0 or later. Different versions of Python can coexist on the same machine. However, you must install them using different versions of PIP. One method of installing different versions of Python is to install Anaconda, which includes PIP. You then use that version of PIP to install a version of Python that supports CAS integration with JupyterHub, SAS Scripting Wrapper for Analytics Transfer (SAS SWAT), and SAS DLPy.
For more information about SAS DLPy, see: https://github.com/sassoftware/python-dlpy/blob/master/README.md.
For more information about SAS SWAT, see https://sassoftware.github.io/python-swat/index.html.