Building Grafana Cloud for next-generation cost-efficiency and scale

Grafana Cloud already has a strong reputation for cost efficiency at scale, but we’re not stopping there. In this session, members of the Grafana Labs engineering and product teams show how Grafana Cloud and the open source databases that power it — Mimir, Loki, Tempo, and Pyroscope — are evolving for the next step-function change in observability data volumes. 

We'll go deep on Mimir, already tested well past 1 billion active series, which is raising the ceiling on ingestion volume and query speed with a new compartmentalized architecture. We'll also give an update on needle-in-the-haystack queries, which we first previewed at GrafanaCon: a new secondary index for Loki that accelerates log queries by shrinking the volume of data that needs to be scanned. We'll share results from early customer deployments as well as what we’ve learned along the way.

Adaptive Telemetry sits on top of this foundation as a second lever for cost efficiency and scale that shrinks data volumes rather than just handling more of them. New capabilities make this more effective than ever before. See a demo of how routing can send telemetry to the storage tier that fits its use case and replay brings colder data back when it matters. Auto-apply and intelligent sampling handle that optimization autonomously, blending in human-in-the-loop control, so you get exactly the telemetry you need, without spending extra time managing it.