Scale Your Metrics with Elasticsearch
summary
Shows concrete scaling choices: a 3‑node Docker Compose cluster with 3 primary shards, zero replicas for the demo, per‑minute indices for the demo ingest, and a coordinating node hashing IDs; recommends shard sizes ~10–50 GB and points out refresh default = 1s. Demonstrates rollover indices to keep shards even, a visual rollups job that aggregates 10s → 5min buckets, and a hot-warm-cold nodes layout; the demo runs and produces rollup documents, though some UI features are still unreleased.
AI-written summary · about
chapters (7)
- 0:00Introduction to Elastic Stack
- 4:42Building blocks of Elasticsearch
- 20:03Architecture for metrics scaling
- 29:56Live demo setup
- 37:08Roll-up features demonstration
- 41:58Index lifecycle management
- 44:28Benchmarking and Q&A
related talks
32mAutoscaling Elasticsearch for Logs on Kubernetes
40mMaking your elastic cluster perform
44mBringing Down the Cost of Your OpenSearch Clusters
35mFrom New Cluster to Insight. Deploying Monitoring and Logging to Kubernetes
18mArchitecting Growth: Scaling Tactics for Prometheus Metrics...
58mTroubleshooting your Elasticsearch cluster like a Support Engineer
41mScaling Observability: From Chaos to Clarity with Open Source Tools
26mGraphite@Scale or How to Store Millions of Metrics per Second