4 Side notes JAX London 2016

Overview

Session: 11th Oct — 11:30 to 12:20 Architectural Resiliency

Notes

• Jeremy Deane has quite a lot of content on line about this subject.

  • Why we need Resilient Software Design
  • What happens when you blame a developer for a problem?
  • Event-driven Microservices slides
  • Architectural Resiliency

• Accumulating innovation debt:

“Innovation debt is the cost that companies incur when they don’t invest in their developers”

• Architecture & design ideas for resilience:
  • Model for threads and failure. “What will we do if … ?”
  • Monitor and measure the results to track what is the actual improvement or deterioration.
    • Meantime to Failure MTTF
    • Meantime to Recovery MTTR
    • Meantime to Deploy MTTD
    • Meantime between failures MTBF
  • Measures do not need to be exact numbers. We are looking for reference points or tendencies.
• Model stateless & asynchronous micro-services
• Do not use distributed transactions
• Implement with CWhatontinuous Delivery
• Have owners for each micro-service.
• Do not blame but motivate people to be accountable for each micro-service.
• Keep the service up to date with continuous improvement. Apply Kanban principles.
• Keep all logging of all services and applications with the same standard format. It reduces time to analyse failures.
• Define recoverability behaviour for:
  • Circuit breaker
  • Throttle
  • Negative testing scenarios
  • Penetration testing
  • Chaos testing: Simian Army
  • Document system external dependencies
  • Practice disaster recovery scenarios (get in shape)
  • Implement test harnesses like
• The thing to do more:
  • For the security features follow the Open Web Application Security Project OWASP.
  • It is paramount to “Validate inputs and standardized outputs”
• Micro-services with event sourcing and Command Query Responsibility Segregation CQRS is a much more resilience system.
  • Follow the 12-factor method to build software-as-a-service apps.
• If you are in the middle of taking a big decision about going or not to implement micro-service architecture, it would help to read this good article. Manage opportunity costs instead of risks
• Overall, to succeed working in a micro-services architecture it is required a Blameless culture in the terms of Etsy’s Winning Secret: Don’t Play The Blame Game!

That was all for this session.

Many thanks for reading, please leave a comment if you have any quality hint.

Keep on testing, better!