• Use Exceptions as the standard way for a script to report execution failure:
    • CONVENTION: LUTFError is used to indicate serious errors
    • CONVENTION: optionally script can return a non-zero value to indicate failure
    • CONVENTION: LUTFError with halt=True is used to halt all subsequent script execution.
    • CONVENTION: halt_on_exception is a global value which can be used for halting execution. Default to False. Can be set to True manually
  • Exceptions should be handled at the top level script run function.
  • Add comments in each test case file which gets pulled out to create a table of test cases.
    • @PRIMARY - is the primary requirement ID
    • @SECONDARY - is the secondary requirement ID
    • @DESCRIPTION - is the test case description
    • File name is the test case ID
  • Clean the system before running the Test
  • Goals for first LUTF release:
    • Run tests
      • Easy way to run a single test
      • Easy way to run a single suite
      • Easy way to run all suites
    • Different LUTF modes working: interactive, batch, daemon
    • Do configuration (LNET)
    • Do clean up
    • Log extraction and storage from each node
      • Wrapper class for logging
        • set logging levels
        • start of logging
        • stop logging
        • Save Log
    • YAML results
    • Existing test cases running properly
    • YAML report to Maloo should work with existing Maloo parsing in order to avoid any changes in Maloo or AT for LUTF integration
      • Need to figure out the format to return to Maloo in order for it to work properly.
  • Add script text to transcripts if scripts are featured in the video
  • Installation: add script that installs all python packages necessary for core LUTF to work
  • Define "agent" as an LUTF entity capable of running scripts no matter local or remote. Base class definition may need to change to include an additional parameter.