5. Observability

• 5.1. Overview
  • 5.1.1. Why Observability Matters
  • 5.1.2. What Observability Includes
  • 5.1.3. How This Chapter Is Organized
• 5.2. Monitoring QuasarDB Clusters
  • 5.2.1. Introduction and Goals
• 5.3. Metrics reference
  • 5.3.1. Cache
  • 5.3.2. Cache - LRU2
  • 5.3.3. Clustering
  • 5.3.4. Environment
  • 5.3.5. Indexes
  • 5.3.6. Network
  • 5.3.7. Performance profiling
  • 5.3.8. Storage
  • 5.3.9. Storage - Async Pipelines
  • 5.3.10. Storage - Backups
  • 5.3.11. Storage - Optimization
  • 5.3.12. Metric Unit Interpretation
• 5.4. Dashboards
  • 5.4.1. Slow Queries Over Time
  • 5.4.2. Error Count and Rate
  • 5.4.3. Persistence Read vs Write Throughput
  • 5.4.4. Client Network In vs Out (Gbit/s)
  • 5.4.5. Session Monitoring
  • 5.4.6. Ingestion Throughput by Bucket Operation
  • 5.4.7. Async Pipelines
  • 5.4.8. Cache Behavior and Evictions
  • 5.4.9. Cache Evictions: Count and Size
  • 5.4.10. Persistence Cache Usage
  • 5.4.11. LRU2 Cache: Page-ins from Disk
  • 5.4.12. LRU2 Cache: Cold Evictions
  • 5.4.13. LRU2 Cache: Promotions to Hot
  • 5.4.14. LRU2 Cache: Hot Evictions
  • 5.4.15. Cache Efficiency: Hot Hits vs Cold Page-ins
• 5.5. Retrieving Statistics from a QuasarDB Cluster
  • 5.5.1. Using QuasarDB Python API
  • 5.5.2. Using qdbsh
• 5.6. Operational Logging
  • 5.6.1. Why Logging Matters
  • 5.6.2. Ingesting and Alerting
  • 5.6.3. Best Practices for Log Management
  • 5.6.4. Key Example Messages
  • 5.6.5. Further Reading
• 5.7. Performance Tracing
• 5.8. User properties
  • 5.8.1. Identifying which connection is pushing data in small increments
  • 5.8.2. Setting user properties from Python API and triggering logging for small increment writes
• 5.9. Troubleshooting Guide
  • 5.9.1. Connection Failures: Operation Timed Out
  • 5.9.2. Connection Failures: Connection Refused
  • 5.9.3. Connection Failures: Client/Server Version Mismatch
  • 5.9.4. Transactional Conflicts: Operation Aborted
  • 5.9.5. Slow Queries or Poor Query Performance
  • 5.9.6. Heavy Disk I/O