All Diagnostics "Failed to Run" state but otherwise able to run jobs... RRS feed

  • Question

  • I am trying to setup a small cluster but I am getting stuck when trying to run the Deployment Environment validator diagnostics. All of the diagnostics (including any other types of diagnostics I try to run) all fail with an odd error message "Internal exception happen when deal with run: Logon failure: unknown user name or bad password."


    The user name that I am using has been added to the local administrators group as well as added as a cluster administrator in HPC Cluster Manager. I am using HPC 2008 R2 Cluster Manager to manage the head node remotely from another workstation on the domain. That seems to work well, as well as allow me to submit trivial jobs to the head node, which run to completion.


    Any ideas as to what might be going on? Thanks!

    Tuesday, August 23, 2011 8:49 PM

All replies

  • I had this error some months ago and i solved this by reinstalling HPC Pack. Odd idea but i think that cluster consistency is OK (also, try to check your cluster with latest LIZARD) but something wrong with HPC Manager.

    Also, you can use Domain Administrator for cluster administration, as far as i know they are mutually connected (HPC Pack authentification <=> ADDS).

    Wednesday, August 24, 2011 6:48 AM
  • *HPC Manager Configuration, sure.
    Wednesday, August 24, 2011 6:49 AM