Updating the BaseOS Image#

This chapter provides information and instructions for updating the BaseOS image with security fixes and critical updates.

The OS image is typically updated “in-place” in BCM using standard package management tools, such as apt for Ubuntu-based images. This method keeps the existing image while installing the latest versions of the software packages.

Alternatively, the OS image can be updated by importing a newer version of the BaseOS image that has been provisioned with the latest software versions. However, this method requires to reapply changes made to the original image by the user and also to reinstall BCM software, such as Slurm or Kubernetes.

Refer to :ref:”updating nodes in bcm” for instructions on assigning the updated image to nodes or node categories.

Important

Before updating the system image of a compute node, refer to Managing Images in BCM for additional information, including the option to clone the image for rolling updates and to rollback changes in case of issues.

Updating the BaseOS Image In-Place from Public Repositories#

The following instructions update a BaseOS image with the latest packages available from public repositories.

  1. Enter the context of the image using the cm-chroot-sw-img command:

    $ cm-chroot-sw-img /cm/images/baseos-image/
    
  2. Verify the current OS release and kernel version. The example is for DGX OS 6.1.0 with kernel version 5.19.0-45-generic.

    $ cat /etc/dgx-release
    
    DGX_NAME="DGX Server"
    DGX_PRETTY_NAME="NVIDIADGX Server"
    DGX_SWBUILD_DATE="2023-08-09-12-30-10"
    DGX_SWBUILD_VERSION="6.1.0"
    DGX_COMMIT_ID="87d8b12"
    DGX_PLATFORM="DGX Server for H100"
    DGX_SERIAL_NUMBER="Not Specified"
    
    $ uname -r
    
    5.19.0-45-generic
    
  3. Run apt update to refresh the local package repository metadata.

    $ apt update
    
  4. You can optionally check the packages that will be upgraded:

    $ apt list --upgradeable
    
  5. Run apt upgrade to update the DGX OS image.

    $ apt upgrade
    
  6. Once apt upgrade is completed, verify the updated DGX OS release.

    $ cat /etc/dgx-release
    
    DGX_NAME="DGX Server"
    DGX_PRETTY_NAME="NVIDIADGX Server"
    DGX_SWBUILD_DATE="2023-08-09-12-30-10"
    DGX_SWBUILD_VERSION="6.1.0"
    DGX_COMMIT_ID="87d8b12"
    DGX_PLATFORM="DGX Server for H100"
    DGX_SERIAL_NUMBER="FMGY9R3"
    DGX_OTA_VERSION="6.3.2"
    DGX_OTA_DATE="Wed May 28 12:16:11 AM PDT 2025"
    
  7. Remove any residual packages.

    $ apt autoremove
    

Less Common Options when Updating an Image#

The following sections provide additional information for less common options.

Pinning Packages#

Pinning packages means to keep a package at the version that is currently installed.

Note

Pinning packages should be used only for very specific instances as it can interfere with package dependencies.

  1. Use the apt-mark hold` command to lock any packages that should not be automatically updated, such as the slurm client.

    $ apt-mark hold slurm-*
    
  2. Use apt-mark unhold to unhold any packages that should be automatically updated, such as the linux kernel and headers. If the MLNX OFED packages were deployed using the BCM repository, these will be on hold and prevent kernel updates.

    $ apt-mark unhold linux-*