Monitoring

Monitoring applications

Key for where information can be obtained: (K)ubernetes, (5) ae5_tools, (S)ystem/OS, (A)pplication (Ac: canned/automatic, Au: user/custom functionality), A, (W)ebserver, [P]roxy in front or your apps, (?)

• Ensure they stay up
  • (Ac) Currently up
  • (K) Uptime (duration and percentage)
  • (P/Ac/?) User failure rate (number and percentage)
• Resource utilization (under/over-utilization)
  • (K) Memory usage
  • (K) CPU usage
  • (K) Number of times memory limits were reached (approximated by k8s restarts)
• Popularity
  • (Ac) Number of visitors
  • (Ac) Duration of visits
  • (Ac) Bounce rate?
  • (Ac) Open sessions
• Responsiveness
  • (Ac) Time to first render
  • (Ac) Responsiveness inside application
• Fine grained tracking/monitoring
  • [Au] Which parts of app are used by visitors?
  • [Au] What error conditions are encountered

Domain conditions

• (Au) Some metric exceeds threshold
• (Au) App raises explicit error condition

Indicators

• Traffic light
• Speedometer/Gauge
• Donut
• KPI (numeric absolute/change)