Elasticsearch Monitoring Without Plugins
Monitor your Elasticsearch clusters without installing any plugins or impacting performance. Get deep insights into query performance, cluster health, shard distribution, and resource usage through wire protocol analysis.
Complete Elasticsearch Observability
Monitor every aspect of your search infrastructure without any modifications
Query Performance Analysis
Track slow queries, search latency, and aggregation performance at the wire level
Cluster Health Monitoring
Real-time cluster status, shard allocation, and node availability without plugins
Index Performance Metrics
Monitor indexing rate, refresh times, merge operations, and document counts
JVM & Resource Tracking
Heap usage, garbage collection, CPU, and thread pool monitoring per node
Shard Distribution Analysis
Track shard balancing, relocations, and hot spots across the cluster
Search Analytics
Query patterns, cache hit rates, and search request distribution insights
Query Performance Insights
Identify Slow Queries Instantly
Search Analytics
- • Query execution time breakdown
- • Search type distribution
- • Aggregation performance
- • Scroll and pagination tracking
Performance Metrics
- • Query latency percentiles
- • Cache hit/miss rates
- • Shard query times
- • Coordinator overhead
# Example slow query detection
{
"query": "match_phrase",
"index": "products",
"took_ms": 342,
"total_shards": 5,
"successful_shards": 5,
"query_body": {
"match_phrase": {
"description": "wireless headphones"
}
}
}Cluster Health Monitoring
Node Metrics
- • CPU and memory usage
- • JVM heap utilization
- • GC frequency and duration
- • Thread pool queues
- • Network I/O rates
Shard Health
- • Primary/replica status
- • Shard allocation
- • Rebalancing progress
- • Unassigned shards
- • Recovery operations
Index Metrics
- • Document count
- • Index size and growth
- • Indexing rate
- • Refresh intervals
- • Merge statistics
Dual Collection Approach
Wire Protocol Analysis
eBPF captures HTTP/JSON traffic to Elasticsearch REST API, extracting query patterns, execution times, and response codes without any overhead.
POST /_search → Track query performanceREST API Metrics
Complementary collection via Elasticsearch APIs provides cluster state, node statistics, and index metrics for complete visibility.
GET /_cluster/health → Cluster statusMonitoring Capabilities
Quick Setup
# Deploy HyperObserve agent docker run -d \ --name hyperobserve \ --pid host \ --network host \ -v /sys/kernel/debug:/sys/kernel/debug:ro \ -e ELASTICSEARCH_URL=http://localhost:9200 \ -e API_KEY=YOUR_API_KEY \ hyperobserve/agent:latest # Works with secured clusters too -e ELASTICSEARCH_USER=elastic \ -e ELASTICSEARCH_PASS=changeme \ -e ELASTICSEARCH_TLS_VERIFY=false
Compatible with Elasticsearch 6.x, 7.x, and 8.x
Monitor Elasticsearch Without Any Plugins
No cluster restarts. No performance impact. Just complete visibility.