• Main INDEX
  • Monthly INDEX
  • PREV
  • NEXT

    User name M. Rvachev

    Log entry time 05:23:01 on December 22,1999

    Entry number 31746

    Followups:

    keyword=High deadtime

    We were taking data stably at 22% deadtime when file copying to MSS started.
    Deadtime started to fluctuate with average of ~40% with low copying speed (~1Mb/s).
    We decided not to change prescale factors and run with the high deadtime. Suffered for 40 minutes. After copying ended, deadtime became ~70%. Restart of coda cured this (kcoda killed 3 processes).

    Since copying speed is very low during taking data, maybe it's more profitable not to run CODA while the copying is done (and increase the speed of copying by this), i.e.
    exchange bad running for a longer time by better running for a shorter time. (This is
    a question).