InfiniBand Cluster Bring-up Procedure
InfiniBand Cluster Bring-up Procedure

MLNX-OS

MLNX-OS can be installed/upgraded using one of the methods below:

Warning

Older versions of the software may require upgrading to one or more intermediate versions prior to upgrading to the latest. Missing an intermediate step may lead to errors.

For further information, see MLNX-OS Release Notes.

  1. Enter Config mode.

    Copy
    Copied!
                

    switch > enable switch # configure terminal switch (config) #

  2. Display the currently available image (.img file).

    Copy
    Copied!
                

    switch (config) # show images Installed images:   Partition 1: <old_image>   Partition 2: <old_image>   Last boot partition: 1 Next boot partition: 1   Images available to be installed: webimage.tbz <old_image>   Serve image files via HTTP/HTTPS: no   No image install currently in progress.   Boot manager password is set.   Image signing: trusted signature always required Admin require signed images: yes   Settings for next boot only: Fallback reboot on configuration failure: yes (default)

  3. Delete the image listed under “Images available to be installed” prior to fetching the new image. Use the command “image delete” for this purpose.

    Copy
    Copied!
                

    switch (config) # image delete <old_image>

    Note

    When deleting an image, it is recommended to delete the file, but not the partition, so as to not overload system resources.

  4. Fetch the new software image.

    Copy
    Copied!
                

    switch (config) # image fetch scp://<username>:<password>@<ip-address>/var/www/html/<new_image> Password (if required): ****** 100.0%[##################################################################]

  5. Display the available images again and verify that the new image now appears under “Images available to be installed”.

    Note

    To recover from image corruption (e.g., due to power interruption), there are two installed images on the system. See the commands “image boot next” and “image boot location” for more information.

    Copy
    Copied!
                

    switch (config) # show images Installed images:   Partition 1: <old_image>   Partition 2: <old_image>   Last boot partition: 1 Next boot partition: 1   Images available to be installed: webimage.tbz <new_image>   Serve image files via HTTP/HTTPS: no   No image install currently in progress.   Boot manager password is set.   Image signing: trusted signature always required Admin require signed images: yes   Settings for next boot only: Fallback reboot on configuration failure: yes (default)

  6. Install the new image.

    Copy
    Copied!
                

    switch (config) # image install <new_image> Step 1 of 4: Verify Image  100.0% [#############################################################] Step 2 of 4: Uncompress Image  100.0% [#############################################################] Step 3 of 4: Create Filesystems  100.0% [#############################################################] Step 4 of 4: Extract Image  100.0% [#############################################################]

    Note

    CPU utilization may go up to 100% during image upgrade.

  7. Have the new image activate during the next boot.

    Copy
    Copied!
                

    switch (config) # image boot next

  8. Run “show images” to review your images.

    Copy
    Copied!
                

    switch (config) # show images Installed images:   Partition 1: <new_image>   Partition 2: <old_image>   Last boot partition: 1 Next boot partition: 1   Images available to be installed: webimage.tbz <new_image>   Serve image files via HTTP/HTTPS: no   No image install currently in progress.   Boot manager password is set.   Image signing: trusted signature always required Admin require signed images: yes   Settings for next boot only: Fallback reboot on configuration failure: yes (default)

  9. Save current configuration.

    Copy
    Copied!
                

    switch (config) # configuration write

  10. Reboot to run the new image.

    Copy
    Copied!
                

    switch (config) # reload Configuration has been modified; save first? [yes] yes Configuration changes saved. Rebooting... switch (config)#

    Note

    After software reboot, the software upgrade will also automatically upgrade the firmware version.

    Note

    On systems with dual management, the software must be upgraded on both the host and the device modules.

For Further information, see MLNX-OS User Manual.

Note

Upgrading MLNX-OS via UFM requires having MFT installed on the UFM server.

To upgrade MLNX-OS via UFM, follow the below steps:

  1. Log into the UFM WEB UI.

  2. Expand the "Managed Elements" and click on Devices.

  3. Identify the switch.

  4. Right-click on the chosen switch.

  5. Click on "Software Upgrade".

    worddavefc74745eb3c1a87daa1d18f3e21d28c-version-1-modificationdate-1716821920033-api-v2.png

  6. Fill the details of the image's location and click Submit.

    worddav6e56582d81ac7b093ff751acfbdef93c-version-1-modificationdate-1716821919577-api-v2.png

  7. Reboot the switch.

    image-2024-4-16_6-21-27-version-1-modificationdate-1716821919077-api-v2.png

© Copyright 2024, NVIDIA. Last updated on May 28, 2024.