NVIDIA UFM Enterprise User Manual v6.12.1
NVIDIA UFM Enterprise User Manual v6.12.1

Installation Notes

Supported NVIDIA Externally Managed Switches

Type

Model

Latest Tested Firmware Version

NDR switches

  • MQM9790

31.2010.4120

HDR switches

  • MQM8790

31.2010.4120

EDR switches

  • SB7790

  • SB7890

15.2008.3118

FDR switches

  • SX6025

  • SX6015

  • SX6005

11.2000.1142

Supported NVIDIA Internally Managed Switches

Type

Model

Latest Tested OS Version

NDR switches

  • MQM9700

MLNX-OS 3.10.4100

HDR switches

  • MQ8700

  • MCS8500

  • TQ8100-HS2F

  • TQ8200-HS2F

MLNX-OS 3.10.4100

EDR switches

  • SB7700

  • SB7780

  • SB7800

  • CS7500

  • CS7510

  • CS7520

MLNX-OS 3.10.3002

FDR switches

  • SX6012

  • SX6018

  • SX6036

  • SX6506

  • SX6512

  • SX6518

  • SX6536

  • SX1012

  • SX6710

  • SX6720

  • SX1700

  • SX1710

MLNX-OS 3.8.1054

Long-haul

  • MTX6100

  • MTX6240

  • MTX6280

MLNX-OS 3.6.8008

InfiniBand-Ethernet Gateways

SX6036G (FDR)

MLNX-OS 3.6.8008

Bare Metal Deployment Requirements

Platform

Type and Version

OS and Kernel(a)

64-bit OS:

  • RedHat 7.9: 3.10.0-1160.el7.x86_64

  • RedHat 8.1: 4.18.0-147.el8.x86_64

  • RedHat 8.2: 4.18.0-193.el8.x86_64

  • RedHat 8.4: 4.18.0-305.el8.x86_64

  • RedHat 8.6: 4.18.0-372.9.1.el8.x86_64

  • RedHat 9.0: 5.14.0-70.13.1.el9_0.x86_64

  • CentOS 7.9: 3.10.0-1160.el7.x86_64

  • CentOS 8 Stream: 5.4.0

  • Ubuntu 18.04: 4.15

  • Ubuntu 20.04: 5.4.0

  • Ubuntu 22.04: 5.4.0

CPU(b)

x86_64

HCAs

  • NVIDIA ConnectX®-4 with Firmware 12.12.xxxx and above

  • NVIDIA ConnectX®-5 with Firmware 16.19.1200 and above

  • NVIDIA ConnectX®-6 with Firmware 20.24.1000 and above

  • NVIDIA ConnectX®-7 with Firmware 28.33.1014 and above

  • NVIDIA Mezzanine Board with Four ConnectX-7 ASICs for Multi-GPU Connectivity (CEDAR) with Firmware 28.36.0394 and above

  • NVIDIA BlueField with Firmware 24.33.900 and above

  • NVIDIA BlueField-2 with Firmware 24.33.900 and above

OFED(c)

  • MLNX_OFED 5.X

Warning

(a) CentOS 8 Stream and RHEL8.4 can be installed without MLNX_OFED; inbox drivers can be used instead.
(b)
CPU requirements refer to resources consumed by UFM. You can also dedicate a subset of cores on a multicore server. For example, 4 cores for UFM on a 16-core server.
(c) For supported HCAs in each MLNX_OFED version, please refer to MLNX_OFED Release Notes.

Warning

From RedHat 9* and onwards, packages with SHA1 signatures are no longer supported. The CONDA package binary is signed with SHA1 signatures and thus, CONDA will not be installed with RedHat 9*.

Two options are available to overcome this.

  1. Recommended Option: Run the following command to install Conda (change gpgcheck from 1 to 0):

    Copy
    Copied!
                

    cat <<EOF > /etc/yum.repos.d/conda.repo [conda] name=Conda baseurl=https://repo.anaconda.com/pkgs/misc/rpmrepo/conda enabled=1 gpgcheck=0 gpgkey=https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc EOF   #install conda yum install conda

  2. Alternative Option: Run the following command to set the RedHat 9* system-wide cryptographic policy to use legacy (less-secured) policy:

    Copy
    Copied!
                

    update-crypto-policies --set LEGACY

    Install Conda as instructed by the UFM installation script.
    After Conda installation, the policy can be set back to default by running the following command:

    Copy
    Copied!
                

    update-crypto-policies --set DEFAULT

Warning

For running SHARP Aggregation Manager within UFM, it is recommended to use MLNX_OFED-5.4.X version or newer.

Warning

Installation of UFM on minimal OS distribution is not supported.

Warning

UFM does not support systems in which NetworkManager service is enabled.

Before installing UFM on RedHat OS, make sure to disable the service.


Docker Installation Requirements

UFM Docker Container is supported on the standard docker environment (engine).

The following operating systems were tested with Docker Container:

Component

Type and Version

Supported OS

  • RHEL7

  • RHEL8

  • Ubuntu18.04

  • Ubuntu20.04

  • Ubuntu22.04

UFM Server Resource Requirements per Cluster Size

Fabric Size

CPU Requirements*

Memory Requirements

Disk Space Requirements

Minimum

Recommended

Up to 1000 nodes

4-core server

4 GB

20 GB

50 GB

1000-5000 nodes

8-core server

16 GB

40 GB

120 GB

5000-10000 nodes

16-core server

32 GB

80 GB

160 GB

Above 10000 nodes

Contact NVIDIA Support

UFM GUI Client Requirements

The platform and GUI requirements are detailed in the following tables:

Platform

Details

Browser

Edge, Internet Explorer, Firefox, Chrome, Opera, Safari

Memory

  • Minimum: 2 GB

  • Recommended: 4 GB

UFM SM Version

Platform

Type and Version

SM

UFM package includes SM version 5.14.0

UFM NVIDIA SHARP Software Version

Platform

Type and Version

NVIDIA® Scalable Hierarchical Aggregation and Reduction Protocol (SHARP)™

UFM package includes NVIDIA SHARP software version 3.2.0

The installer detects versions previously installed on the machine and prompts you to run a clean install of the new version or to upgrade while keeping user data and configuration unchanged.

The upgrade from previous versions maintains the existing database and configuration, allowing a seamless upgrade process.

Note

Upgrading UFM Enterprise software version is supported up to two previous GA software versions (GA -1 or -2).
For example, if you wish to upgrade to UFM Enterprise v6.11.0, it is possible to do so only from UFM Enterprise v6.9.0 or v6.10.0.

For detailed installation and upgrade instructions, refer to the UFM Quick Start Guide or the UFM User Manual.

Warning

Due to a possible conflict, SM and MFT installed by the MLNX_OFED must be uninstalled. The installation procedure will detect and print all MLNX_OFED packages that must be removed.

Warning

It is recommended to upgrade to the latest UFM version from the last 2 GA releases that came before it. Upgrading from older UFM versions may result in failures.

© Copyright 2023, NVIDIA. Last updated on Sep 5, 2023.