Lately, there’s been a lot of talk around Observability 2.0 and a broader transformation from the three (or four, depending on who you ask) pillars of observability towards a unified source of truth: arbitrarily wide, structured log events.

If we boil this down, we’re right back to where we started — kind of. In the beginning, there were logs, and now we’ve come full circle, with wider logs (structured, high-dimensionality time series events) and more of them than ever before (plus high cardinality). So, how did we get here? And where exactly are we headed?