'total_gaudi_cards' is undefined. 'total_gaudi_cards' is undefined

[yaoboyuan]

[Environment]
OS: Ubuntu 22.04.5 LTS
Kernel: 6.8.0-40-generic
Hardware: Gaudi3 HL-325
Gaudi driver: 1.21.3-57/
Code version: release-1.4.0

[Issue]
Encounter below error when running ansible-playbook -K playbooks/infrastructure.yaml --tags configure,install -i inventory/test-cluster/inventory.ini -e @inventory/test-cluster/config.yaml

TASK [gaudi_operator : Print total number of Gaudi cards] *******************************************************************************************************************
週五 22 八月 2025  14:03:21 +0800 (0:00:00.123)       0:18:20.639 *****************
fatal: [smc]: FAILED! =>
  msg: |-
    The task includes an option with an undefined variable. The error was: 'total_gaudi_cards' is undefined. 'total_gaudi_cards' is undefined

    The error appears to be in '/home/brian/Enterprise-RAG/deployment/roles/infrastructure/gaudi_operator/tasks/main.yaml': line 221, column 3, but may
    be elsewhere in the file depending on the exact syntax problem.

    The offending line appears to be:


    - name: Print total number of Gaudi cards
      ^ here

[yaoboyuan]

I also try gaudi driver 1.21.4-3/
But still the same issue

[cepera]

The issue happens because the playbooks are not being able to install the gaudi operator correctly (most likely due to slow internet connectivity), the timeout variably from the inventory doesn't seem to be propagated to the gaudi operator installation tasks.
We'll be working on a fix to ensure the timeout configuration is correctly applied during operator installation.

As a temporary solution, please wait until all Gaudi operator pods are in a ready state (after the The error was: 'total_gaudi_cards' is undefined error), then re-run the playbook with the configure,install tags.

[yaoboyuan]

Yes, the temporary solution works