20 - 21 SEPTEMBER 2017 / Stockholmsmässan

DevOps and Open Source Theatre

Wed 20th Sep 14:50 to 15:20

Don't Let your Container Applications Run Away: Best Practices Implementing Monitoring, Alerting, Daily Live Site Triage and Root Cause Diagnostics

Over the past several years, with proliferation of micro-services and now, increasingly, micro-services running in containers, monitoring has become progressively complex and smart, low-overhead, low-storage cost monitoring has become as critical as never before. Among the unique challenges that DevOps are facing are (1) Additional complexity that containers are adding with their transient lifecycle; (2) Sheer volume of data being collected; (3) Learning to effectively apply new innovative monitoring approaches such as adaptive sampling, consolidated alerting, predictive analytics. In this talk I will cover best monitoring practices, patterns and anti-patterns and monitoring solutions available for different technologies. As part of this talk, we’ll walk through setting up end to end monitoring dashboard for a medium-size complexity micro-service application running in Docker, tracing transactions across multiple services and diagnosing an issue.

What you will take away from this session

  • Why monitoring is critical for containerized applications and why it represents unique challenges
  • How to effectively plan daily live site triage, diagnostics and root cause analysis for a microservice application with services owned by multiple teams
  • What metrics to alert on, what metrics to observe daily and which ones to collect for deeper analysis
  • How to deal with large volumes of monitoring data and find the right balance to preserve some details for transactions for investigations and not to overpay for monitoring storage


Photo Speaker Name Profile
Alex Bulankou Alex Bulankou View Profile