Versions Compared

    Key

    • This line was added.
    • This line was removed.
    • Formatting was changed.
    Comment: Migrated to Confluence 5.3
    Info

    New in Qube 6.5

    Another new feature in Qube 6.5 is tracking "cumulative time", and this is stored as cpu-seconds.  This will show in WranglerView and ArtistView as "CPU-minutes".

    If a job instance or frame runs for 60 seconds, and has been allocated 3 slots, the "timecumulative" value is set to 180 when the instance or frame is complete/failed/killed. If the same job or instance is retried and is allocated 8 slots this time around, and runs for 30 seconds, the "cumulative time" is increased by (8 * 30) 240 cpu-seconds, for a new value of 180 + 240 = 420 cpu-seconds.

    We feel that cpu-minutes is a much more accurate metric for "how much of the farm did I get?" that trying to track cores or processors.

    See Also

    Job Reservations