Lightning Talk
summary
A Prometheus-compatible time-series database that replaces Prometheus long-term storage while keeping PromQL and remote_write compatibility, with components VMInsert, VMStorage and VMSelect and a VM Agent for cluster-wide scraping. Claims include 5× less RAM and 7× less disk space than Prometheus, much lower disk I/O (quoted 95 MB vs ~15 MB), higher write throughput in benchmarks, and single-node clusters handling very large ingestion volumes without OS spikes. Deployment is a config change in Prometheus to redirect remote_write and swap the TSDB.
AI-written summary · about
chapters (9)
- 0:00Intro
- 0:28Raghu Reddy
- 1:02ABOUT
- 1:37VM CLUSTER
- 2:49VM AGENT
- 3:19WHY VICTORIA METRICS
- 3:59Cool Benchmark Reac
- 4:16Cool Benchmark Write
- 4:42FEATURES AND BENEFITS
related talks
04mLightning Talk: TSDB: The Engine behind Prometheus - Goutham Veeramachaneni, IIT Hyderabad
30mTSDB: The Past, Present and the Future - Goutham Veeramachaneni, IIT Hyderabad
46mPrometheus as a Time Series Database
05mLightning Talk: Transactional Remote-write and Why You Should Care
39mBuild a High Performance Remote Storage for Prometheus with Unlimited Time Series
55mAdding ++ to Prometheus
05mMigrating Prometheus data between different storage systems
40mCost-Effective, Dense, and Performant Prometheus Storage via QLC