This release brings a better storage that will be able to build bigger blocks of data. Next up we're working on reducing memory usage of ingesters.
We've seen speed up of queries across 200 profiled pods. 🔥
There's also couple of interesting bugfixes and improvements see the list 👇
Changelog
- Flush profiles to disk per row groups by @simonswine in #486
- Add observability using Loki/Tempo and Grafana Agent by @simonswine in #495
- Add a debug image with phlare running through dlv by @simonswine in #511
- Fix regressions on the new ingress flush to disk handling. by @simonswine in #509
- Fix cutting by block size and adjust limits by @simonswine in #514
- Use the correct API in readyHandler by @hi-rustin in #516
- Increase ingesters default limits by @cyriltovena in #518
New Contributors
- @hi-rustin made their first contribution in #515
Full Changelog: v0.2.0...v0.3.0
As always, feedbacks are more than welcome, feel free to open issues/discussions.
You can reach out to the team using:
Docker Images
docker pull docker.io/grafana/phlare:0.3.0