NVIDIA UFM Enterprise User Manual v6.18.0
NVIDIA UFM Enterprise User Manual v6.18.0

Upgrading UFM on Docker Container

Note

Upgrade the UFM container based on the existing UFM configuration files that are mounted on the server. It is important to use that same directory as a volume for the UFM installation command.

In the below example /opt/ufm_files is used.

  1. Stop the UFM Enterprise service. Run:

    Copy
    Copied!
                

    systemctl stop ufm-enterprise

  2. Remove the existing docker image. Run:

    Copy
    Copied!
                

    docker rmi mellanox/ufm-enterprise:latest

  3. Load the new UFM Enterprise docker image. Run:

    Copy
    Copied!
                

    docker pull mellanox/ufm-enterprise:latest

  4. Run the docker upgrade command:

    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/:/opt/ufm/shared_config_files/ \ mellanox/ufm-enterprise:latest --upgrade

  5. Reload system manager configuration:

    Copy
    Copied!
                

    systemctl daemon-reload

  6. Start UFM Enterprise service:

    Copy
    Copied!
                

    systemctl start ufm-enterprise

Note

As of UFM version 6.14.0, UFM upgrade on HA supports in-service upgrade, meaning UFM can continue running during the steps of the upgrade, and there is no need to stop UFM before the upgrade (although this is also supported).

Upgrading the UFM Enterprise Package

  1. Remove the old docker image from the Standby server. Run:

    Copy
    Copied!
                

    Stand-by# docker rmi mellanox/ufm-enterprise:latest

  2. Pull the new UFM Enterprise docker image on the Standby server. Run:

    Copy
    Copied!
                

    docker pull mellanox/ufm-enterprise:latest

    Note

    At this stage, the UFM container has been updated with the latest code. The UFM data, however, will be updated during the next UFM run.

  3. Perform a failover to start UFM on the upgraded node. On the Master node, run:

    Copy
    Copied!
                

    ufm_ha_cluster failover

    Note

    When UFM starts, it will automatically update the UFM configuration.

  4. Repeat steps 1-2 on the un-upgraded node (previous Master node).

Upgrading the UFM HA Package

  1. On both servers, download and extract the latest UFM HA package. Run:

    Copy
    Copied!
                

    wget https://www.mellanox.com/downloads/UFM/ufm_ha_5.6.0-4.tgz

    For Sha256:

    Copy
    Copied!
                

    wget https://download.nvidia.com/ufm/ufm_ha/5.6.0/ufm_ha_5.6.0-4.sha256

  2. Stop the UFM HA cluster, run the following command on the Master server:

    Copy
    Copied!
                

    ufm_ha_cluster stop

  3. On both the Master and Standby servers, execute the upgrade command from within the extracted HA package. Ensure you run it first on the Standby server, then on the Master server:

    Copy
    Copied!
                

    ./install.sh --upgrade

  4. Start the UFM HA cluster by running this command on the Master server:

    Copy
    Copied!
                

    ufm_ha_cluster start

  5. Run the following command to verify that the UFM HA cluster is up and running:

    Copy
    Copied!
                

    ufm_ha_cluster status

© Copyright 2024, NVIDIA. Last updated on Aug 22, 2024.