This quote from Charity Majors is probably the best summary of the current state of observability in the tech industry - a total, mass confusion. Everyone is confused. What is a trace? What is a span? Is log line a span? Do I need traces if I have logs? Why I need traces if I have great metrics? The list of questions like these goes on. Charity - together with other great folks from observability system called
Scuba kinda sounds like what Loki and other “modern” log aggregation tools do: A message field and lots of metadata that is efficiently stored so queries are fast.
I do understand the poster’s frustration with the “three pillars”, and agree that Open telemetry feels over-engineered at times (seriously… what’s “baggage”?).
But the three pillars really do all have a place! While we absolutely can generate stats from logs, the storage improvements alone from leveraging a TSDB are worth it as we scale. And tracing gives incredibly unique insight into the path a request takes through our systems.
Scuba kinda sounds like what Loki and other “modern” log aggregation tools do: A message field and lots of metadata that is efficiently stored so queries are fast. I do understand the poster’s frustration with the “three pillars”, and agree that Open telemetry feels over-engineered at times (seriously… what’s “baggage”?). But the three pillars really do all have a place! While we absolutely can generate stats from logs, the storage improvements alone from leveraging a TSDB are worth it as we scale. And tracing gives incredibly unique insight into the path a request takes through our systems.