NVIDIA UFM Enterprise User Manual v6.15.16 LTS (2023 LTS U6)
NVIDIA Docs Hub Homepage  NVIDIA Networking  Networking Software  Management Software  NVIDIA UFM Enterprise User Manual v6.15.16 LTS (2023 LTS U6)  UFM Docker Installation

On This Page

UFM Docker Installation

General Prerequisites

  • MLNX_OFED must be installed on the server that will run UFM Docker

  • For UFM to work, you must have an InfiniBand port configured with an IP address and in "up" state.

    Note

    For InfiniBand support, please refer to NVIDIA Inbox Drivers , or Mellanox_OFED guides.

  • Make sure to stop the following services before running UFM Docker container, as it utilizes the same default ports that they do: Pacemaker, httpd, OpenSM, and Carbon.

  • If firewall is running on the host, please make sure to add an allow rule for UFM used ports (listed below):

    Note

    If the default ports used by UFM are changed in UFM configuration files, make sure to open the modified ports on the host firewall.

    • 80 (TCP) and 443 (TPC) are used by WS clients (Apache Web Server)

    • 8000 (UDP) is used by the UFM server to listen for REST API requests (redirected by Apache web server)

    • 6306 (UDP) is used for multicast request communication with the latest UFM Agents

    • 8005 (UDP) is used as a UFM monitoring listening port

    • 8888 (TCP) is used by DRBD to communicate between the UFM Primary and Standby servers

    • 2022 (TCP) is used for SSH

Prerequisites for Upgrading UFM Docker Container

  • Supported versions for upgrade are UFM v.6.7.0 and above.

  • UFM files directory from previous container version mounted on the host.

Step 1: Loading UFM Docker Image

To load the UFM docker image, pull the latest image from docker hub:

Copy
Copied!
            

            
docker pull mellanox/ufm-enterprise:latest

Note

You can see full usage screen for ufm-installation by running the container with -h or -help flag:

Copy
Copied!
            

            
docker run --rm mellanox/ufm-enterprise-installer:latest -h

Step 2: Installing UFM Docker

Installation Command Usage

Copy
Copied!
            

            
docker run -it --name=ufm_installer --rm \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /etc/systemd/system/:/etc/systemd_files/ \
-v [UFM_FILES_DIRECTORY]:/installation/ufm_files/ \
-v [LICENSE_DIRECTORY]:/installation/ufm_licenses/ \
mellanox/ufm-enterprise:latest \
--install [OPTIONS]

Modify the variables in the installation command as follows:

  • [UFM_FILES_DIRECTORY]: A directory on the host to mount UFM configuration files.

    Note

    UFM_FILES_DIRECTORY must have read/write permissions for other users because UFM needs write data during runtime.

    Note

    Example: If you want UFM files on the host to be under /opt/ufm_files/ you must set this volume to be: -v /opt/ufm_files/:/installation/ufm_files/

  • [UFM_LICENSES_DIR]: UFM license file or files location.

    Note

    Example: If your license file or files are located under /downloads/ufm_license_files/ then you must set this volume to be -v /downloads/ufm_license_files/:/installation/ufm_licenses/

  • [OPTIONS]: UFM installation options. For more details see the table below.

Command Options

Flag

Description

Default Value

-f | --fabric-interface

IB fabric interface name.

ib0

-g | --mgmt-interface

Management interface name.

eth0

-h | --help

Show help

N/A

Installation Modes

UFM Enterprise installer supports several deployment modes:

Stand Alone (SA) Installation

  1. Create a directory on the host to mount and sync UFM Enterprise files with read/write permissions. For example: / opt/ufm_files/.

  2. Copy only your UFM license file(s) to a temporary directory which we’re going to use in the installation command. For example: /tmp/license_file/

  3. Run the UFM installation command according to the following example which will also configure UFM fabric interface to be ib1:

    Copy
    Copied!
                
    
            
    docker run -it --name=ufm_installer --rm \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /etc/systemd/system/:/etc/systemd_files/ \
-v /opt/ufm_files/:/installation/ufm_files/ \
-v /tmp/license_file/:/installation/ufm_licenses/ \
mellanox/ufm-enterprise:latest \
--install \
--fabric-interface ib1

    Note

    The values below can be updated in the command to your needs:

    • /opt/ufm_files/

    • /tmp/license_file/

    • For example, if you want UFM files to be mounted in another location on your server, create that directory and replace the path in the command.

  4. Reload system

    Copy
    Copied!
                
    
            
    systemctl daemon-reload

  5. To Start UFM Enterprise service run:

    Copy
    Copied!
                
    
            
    systemctl start ufm-enterprise

High Availability

This solution requires downloading and installing a new package released with UFM v6.9.0 called “UFM-HA”.

Pre-deployments requirements

  • Install pacemaker, pcs, and drbd-utils on both servers

  • A partition for DRBD on each server (with the same name on both servers) such as /dev/sdd1. Recommended partition size is 10-20 GB, otherwise DRBD sync will take a long time to complete.

  • CLI command hostname -i must return the IP address of the management interface used for pacemaker sync correctly (update /etc/hosts/ file with machine IP)

  • Create the directory on each server under /opt/ufm_files/ with read/write permissions on each server. This directory will be used by UFM to mount UFM files, and it will be synced by DRBD.

Installing UFM Containers

On the main server, install UFM Enterprise container with the command below:

Copy
Copied!
            

            
docker run -it --name=ufm_installer --rm \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /etc/systemd/system/:/etc/systemd_files/ \
-v /opt/ufm_files/:/installation/ufm_files/ \
-v /tmp/license_file/:/installation/ufm_licenses/ \
mellanox/ufm-enterprise:latest \
--install

On each the standby (secondary) server, install UFM Enterprise container like the following example with the command below:

Copy
Copied!
            

            
docker run -it --name=ufm_installer --rm \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /etc/systemd/system/:/etc/systemd_files/ \
-v /opt/ufm_files/:/installation/ufm_files/ \
mellanox/ufm-enterprise:latest \
--install


Downloading UFM HA Package

Download UFM-HA package on both servers using the following command:

Copy
Copied!
            

            
wget https://www.mellanox.com/downloads/UFM/ufm6.9_ha_2.0.0-23.tgz


Installing UFM HA Package

  1. Extract the downloaded UFM-HA package on both servers under /tmp/

  2. Go to the extracted directory /tmp/ufm_ha_XXX and run the installation script:

    Copy
    Copied!
                
    
            
    ./install.sh -f /opt/ufm_files/ -p /dev/sda5

    Option

    Description

    -f

    UFM enterprise files directory

    -p

    Partition name for DRBD

  3. Configure HA from the main server using the following command:

    Copy
    Copied!
                
    
            
    configure_ha_nodes.sh \
--cluster-password 123456 \
--main-hostname ufm-host01 \
--main-ip 192.168.10.1 \
--main-sync-interface enp2s0f0 \
--standby-hostname ufm-host02 \
--standby-ip 192.168.10.2 \
--standby-sync-interface enp2s0f0 \
--virtual-ip 192.168.10.5
    Note

    configure_ha_nodes.sh will require SSH connection to the standby server. If SSH is not configured then you will be prompted to enter the password during configuration runtime.

    Option

    Description

    --cluster-password

    UFM HA cluster password for authentication by pacemaker.

    --main-hostname

    Master (main) server hostname

    --main-ip

    Master (main) server IP address

    --main-sync-interface

    Port name (interface) on master (main) server that will be used in DRBD sync

    --standby-hostname

    Standby server hostname

    --standby-ip

    Standby server IP address

    --standby-sync-interface

    Port name (interface) on standby server that will be used in DRBD sync

    --virtual-ip

    UFM HA cluster Virtual IP

  4. Wait until after configuration for DRBD sync to finish depending on the size of your partition. You can check the status of DRBD sync by running the following command:

    Copy
    Copied!
                
    
            
     drbd-overview

  5. To start UFM HA cluster:

    Copy
    Copied!
                
    
            
     ufm_ha_cluster start

  6. To check UFM HA cluster status:

    Copy
    Copied!
                
    
            
    ufm_ha_cluster status

  7. To stop UFM HA cluster:

    Copy
    Copied!
                
    
            
    ufm_ha_cluster stop

  8. To uninstall UFM HA, first stop the cluster and then the ufm_ha uninstallation script as follows:

    Copy
    Copied!
                
    
            
      /opt/ufm/ufm_ha/uninstall_ha.sh

Upgrading From Existing UFM Container

Note

The installer will upgrade the UFM container based on the existing UFM configuration files that are mounted on your server, so make sure to use that same directory as a volume for the UFM installation command.

Upgrade Container Command

Run the installer like the following example (Assuming your old UFM container uses /opt/ufm_files/ to keep UFM files mounted on the host)

Copy
Copied!
            

            
docker run -it --name=ufm_installer --rm \
-v /var/run/docker.sock:/var/run/docker.sock \ 
-v /opt/ufm_files/:/opt/ufm/shared_config_files/ \
mellanox/ufm-enterprise:latest \
--upgrade

Logging Into UFM Web UI

In order to open UFM Web UI, open the following URL in your browser: https://[SERVER_IP]/ufm/.

Note

The default credentials are:

  • Username: admin

  • Password: 123456
© Copyright 2026, NVIDIA. Last updated on Jan 6, 2026
content here