Skip to content
Dec 20, 2012


v1.1 Release Notes

* improved robustness

  - option --retries: retry failed jobs
  - option --strikes: strike out workers that rack up N consecutive failures

  - watchdog process
      kills zombie workers
      destroys leftover instances

  - autoresume pending jobs
  - retry failed ssh pings

* SSH authentication: --ssh-identity option

* Convenience: --retry=session-id option

  Sets state of all failed jobs to pending, then resumes session

* Better reports (logalyzer)

  - better subject line
  - calculate cost/efficiency
  - show last output from failures

* Better logging

  - verboseness: log/print more information on what is happening, especially
    instance boot/destruction

  - filter progress bars and other noisy \r crap

* Many bugfixes and refactorings
Aug 11, 2011


v1.0 Release Notes

* first full implementation.
Something went wrong with that request. Please try again.