DevOps Classroom notes 31/Jan/2024

Tale of Food delivery application

• Architecture
  
• Possible failures:
  • Connectivity issues
  • Hardware issues
  • OS issues
  • Web server/app server failures
  • applications
  • performance bottlenecks
    • cpu
    • RAM
    • disk
• Our idea is to ensure we have low MTTR (Mean time to recover) and high MTBF (Mean Time between failures) and MTTF (Mean time to fail)
• To do this we need to monitor
  • heart beat or alive
    • servers
    • applications
  • performance monitoring
    • cpu utilization
    • free disk space
    • free RAM
  • log monitoring
  • Trace monitoring / Profiling
• Centralized Monitoring: All the metrics, logs, trace should be centralized to analyse failure or predict failures without need to login into each server.

Monitoring Types

• There are two types of monitoring
  • Server Monitoring
  • Application Monitoring (logs, traces, metrics)
• Softwares to help here
  • Server Monitoring:
    • Nagios
    • Zabbix
  • Application Monitoring
    • New Relic
    • App dynamics
    • splunk
    • prometheus
  • Cloud Monitoring
    • AWS Cloudwatch
    • Azure Monitor
• Tools we will understand
  • Elastic Stack
  • Prometheus, Grafana (kubernetes)
• Technologies to watch out
  • eBPF
  • Open Telemetry

Methodologies

• ITIL
• Continuous Monitoring
• SRE