• Julien Bisconti

    Julien Bisconti

    6 months ago
    Hi team, we are trying to compute the overhead time of the platform. Meaning:
    1 - (All workflows transition time / all workflows duration)
    We use this PromQL query:
    1 - 
      ( flyte:propeller:all:workflow:completion_latency_unlabeled_ms_sum
        + flyte:propeller:all:node:transition_latency_unlabeled_ms_sum
        + flyte:propeller:all:node:queueing_latency_unlabeled_ms_sum
        + flyte:propeller:all:workflow:acceptance_latency_unlabeled_ms_sum)
      / (flyte:propeller:all:workflow:failure_duration_unlabeled_ms_sum + flyte:propeller:all:workflow:success_duration_unlabeled_ms_sum)
    The numbers we get seem a bit off ( around 50%). Is the PromQL query correct? What do you recommend we use to measure that overhead time ?
  • Ketan (kumare3)

    Ketan (kumare3)

    6 months ago
    @Julien Bisconti overhead is always compared to the total runtime and that depends on the workload
  • Also happy to connect and help optimize knobs
  • Julien Bisconti

    Julien Bisconti

    6 months ago
    yes please. We are struggling to understand the numbers we see
  • What is the metrics for total runtime of a workflow ?
  • Should we do a video call ? Are you available ? https://meet.google.com/xxh-cbzt-kkd @Ketan (kumare3)
  • Ketan (kumare3)

    Ketan (kumare3)

    6 months ago
    We can, a little too early for me right now
  • Cc @Dan Rammer (hamersaw) if you are around
  • But it's a holiday in us today
  • Let me dm