Today, mlive doesn't really look at request latency, but high latency can be as bad as downtime. It would be useful if it reported some basic stats about request latency over each interval that it report (e.g., min, median, 99th percentile, max for each of read and write)
Today, mlive doesn't really look at request latency, but high latency can be as bad as downtime. It would be useful if it reported some basic stats about request latency over each interval that it report (e.g., min, median, 99th percentile, max for each of read and write)