Words: Chaos Engineering

"Chaos Engineering" is a rather cool IT discipline, but I haven't had the privilege of implementing it—perhaps because the systems I work with are never complex enough to require such a high degree of redundancy. Essentially, chaos engineering involves testing a system by introducing controlled faults or failures to identify its weaknesses and ensure it can handle unexpected disruptions.

Examples of Chaos Engineering:

  • Killing servers or nodes in a High Availability configuration to see if the service continues to run as expected.
  • Introducing latency between servers/nodes (in a microservices environment) to observe how the system handles slow communication.

Why It’s Important:

  • Proactive Resilience: It helps you uncover weaknesses before they escalate into actual incidents, enabling teams to build more resilient systems.
  • Improved Incident Response: By simulating real-world failures, teams gain valuable experience in handling outages, making them better equipped to respond swiftly and effectively.
  • Confidence in Scalability: As your system grows, chaos engineering allows you to test its ability to scale without hitting unforeseen issues.

Think of it as a 'fire drill' for your system—testing its ability to recover from failures under controlled circumstances. It serves as a great complement to your Disaster Recovery Process, especially if your organization is ISO 27001:2022 certified.

A notable open-source tool used for this is Chaos Monkey by Netflix, which randomly shuts down servers in production to test resilience. You can explore more here: Chaos Monkey by Netflix.

If you're curious about trying this out, start small by experimenting in a development or staging environment before running tests in production.

Comments

Post a Comment

Popular posts from this blog

Words: You Aren't Gonna Need It (YAGNI)

Another daily muse. It’s not surprising that our technical team continues to discover unused functions within a particular feature library. Although the names of these functions may sound promising, the logic inside is often outdated, as they haven’t been refactored to align with the current context of the feature. Some of these functions were written ages ago. This is where I need to borrow a concept from my university days, when my mates and I often applied Extreme Programming (XP) principles in our software projects. The core idea: Don’t add functionality until it’s actually needed. This is essentially the You Aren't Gonna Need It (YAGNI) philosophy. To quote Ron Jeffries, a co-founder of XP, as taken from Wikipedia: "Always implement things when you actually need them, never when you just foresee that you [will] need them."
Read more

Words: Domain-Driven Development

Today, I stumbled upon a simple read on Reddit that serves as a reminder for both new and seasoned programmers about the importance of Domain-Driven Development (DDD). The post shared this article by Google: Write Change-Resilient Code with Domain-Driven Design . While it just touches on DDD, it's good enough to raise awareness. In case DDD is new to you, here’s a brief description from Wikipedia: Domain-Driven Development is a major software design approach that focuses on modeling software to match a domain based on input from that domain’s experts. DDD opposes the idea of having a single, unified model. Instead, it divides a large system into bounded contexts, each with its own model. For a more in-depth exploration of DDD, here are some useful articles: DDD 101: The 5-Minute Tour Domain-Driven Design (DDD) From my perspective, the DDD concept allows for simpler collaboration between technical teams and domain experts through the use of a consistent language. It also reduces th...
Read more

Finds: "Microservices vs. Monoliths" - A Reddit Discussion

Here’s another interesting discussion I stumbled upon on Reddit. The debate between "Microservices vs. Monoliths" is a frequent one, but in my opinion, the context matters—both approaches have their own merits depending on the situation. I plan to write a more in-depth entry on this topic in the near future. Enjoy reading:  Microservices vs. Monoliths: Why Startups Are Getting "Nano-Services" All Wrong
Read more