Observability

• 1. Overview
  • 1.1. Why Observability Matters
  • 1.2. What Observability Includes
  • 1.3. How This Chapter Is Organized
• 2. Monitoring QuasarDB Clusters
  • 2.1. Introduction and Goals
• 3. Metrics reference
  • 3.1. Cache
  • 3.2. Cache - LRU2
  • 3.3. Clustering
  • 3.4. Environment
  • 3.5. Indexes
  • 3.6. Network
  • 3.7. Performance profiling
  • 3.8. Storage
  • 3.9. Storage - Async Pipelines
  • 3.10. Storage - Backups
  • 3.11. Storage - Optimization
  • 3.12. Metric Unit Interpretation
• 4. Dashboards
  • 4.1. Slow Queries Over Time
  • 4.2. Error Count and Rate
  • 4.3. Persistence Read vs Write Throughput
  • 4.4. Client Network In vs Out (Gbit/s)
  • 4.5. Session Monitoring
  • 4.6. Ingestion Throughput by Bucket Operation
  • 4.7. Async Pipelines
  • 4.8. Cache Behavior and Evictions
  • 4.9. Cache Evictions: Count and Size
  • 4.10. Persistence Cache Usage
  • 4.11. LRU2 Cache: Page-ins from Disk
  • 4.12. LRU2 Cache: Cold Evictions
  • 4.13. LRU2 Cache: Promotions to Hot
  • 4.14. LRU2 Cache: Hot Evictions
  • 4.15. Cache Efficiency: Hot Hits vs Cold Page-ins
• 5. Retrieving Statistics from a QuasarDB Cluster
  • 5.1. Using QuasarDB Python API
  • 5.2. Using qdbsh
• 6. Operational Logging
  • 6.1. Why Logging Matters
  • 6.2. Ingesting and Alerting
  • 6.3. Best Practices for Log Management
  • 6.4. Key Example Messages
  • 6.5. Further Reading
• 7. Performance Tracing
• 8. User properties
  • 8.1. Identifying which connection is pushing data in small increments
  • 8.2. Setting user properties from Python API and triggering logging for small increment writes
• 9. Troubleshooting Guide
  • 9.1. Connection Failures: Operation Timed Out
  • 9.2. Connection Failures: Connection Refused
  • 9.3. Connection Failures: Client/Server Version Mismatch
  • 9.4. Transactional Conflicts: Operation Aborted
  • 9.5. Slow Queries or Poor Query Performance
  • 9.6. Heavy Disk I/O