• Main INDEX
  • Monthly INDEX
  • PREV
  • NEXT

    User name R. Michaels

    Log entry time 04:46:57 on June25,2000

    Entry number 44858

    Followups:

    keyword=more about ET losses

    It looks like the various scripts that insert files, EPICS, or scalers
    can interfere with each other and have a probabilty of order 0.5%
    to lose an insertion. I tried to reduce the overlap time by explicitly
    detaching from ET right after insertion, which seems to have cut down
    on one type of interference but created another of comparable frequency
    and same symptom. Instead of failing to open a connection because of a
    lingering attachment, it seems the explicit detachment can kill another process.
    But the analyzers like ESPACE do NOT interfere because they attach an
    independent way. Oh well, 0.5% loss is not a big problem.