Sly Reliability - Part 2 of Bad Observability

Too much detail is hindering us from seeing the pattern. An average in a percentile or any kind of aggregate can hide those patterns. So I've given some examples, but it's not the examples that matter. The important message is that we need to understand the metrics that we're using and what they mean and what they don't mean. Because otherwise we're going to be led to the wrong conclusion and we might make decisions which cause us problems"