Making your elastic cluster perform
summary
Practical, checklist-style guidance for running Elasticsearch: set JVM heap <=30g and give the OS the rest, prefer many CPU cores and SSDs; configure cluster name and min_master_nodes as a majority; design mappings with multi-fields and avoid dynamic surprises (use aliases for zero-downtime reindex); keep shards modest (rule of thumb shard ~50GB), start small and test; speed up bulk loads by using the bulk API, disabling refresh and disable replicas during bulk then re-enable; use the Profile API, Marvel/Kibana, slowlog and metrics (GC, fielddata, fd) to find hotspots.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 0:48Why use elastic search
- 1:43Getting started
- 2:31Overview
- 3:19Installation configuration
- 7:35How many notes
- 18:43Design your mapping
- 25:03Indexing documents
- 27:17Optimize during indexing
- 28:56Bulk indexing
- 29:49Query faster
- 31:05Request body query
- 32:13Query context
- 35:08Aggregations
- 35:38Dog Values
- 37:05Notes
- 38:57Dynamic logging
related talks
16mTop Elasticsearch tips
44mBringing Down the Cost of Your OpenSearch Clusters
01h 01mBest Practices With Elasticsearch
01h 01mElasticsearch Do's, Don'ts and Pro-Tips
45mMonitoring Your OpenSearch Cluster Effectively
29mHow (not) to Scale Elasticsearch for Data Analytics!
58mTroubleshooting your Elasticsearch cluster like a Support Engineer
41mTips and Tricks to Scale Elasticsearch for 1M RPMs and Beyond