Sparse High-resolution Histograms in the Prometheus TSDB
CNCF [Cloud Native Computing Foundation] via YouTube
Overview
Explore the experimental support for sparse high-resolution histograms in the Prometheus Time Series Database (TSDB) through this informative conference talk. Delve into the latest improvements, including storage and indexing space savings, as well as benchmarks for storing and retrieving the new histograms. Gain insights into the design process, implementation challenges, and interesting discoveries made during development. Learn about sparse buckets, histogram chunks, and the data structure used. Examine benchmark results and understand the limitations and future work planned for this feature. Perfect for developers and end-users of Kubernetes, Prometheus, and other CNCF-hosted projects looking to enhance their understanding of time series data management.
Syllabus
Introduction
Recap
Improvements
Design Dock
Hackathon
Heatmap
Overview
Sparse Buckets
Histograms Chunk
Test Setup
Data Structure
Data Snapshot
Benchmark Results
Conclusion
Partitioning
Limitations
Future work
Outro
Taught by
CNCF [Cloud Native Computing Foundation]