@opsmatters.comThe latest update for #GremlinInc includes "Why #AI development creates a reliability blind spot for humans, and what to do about it" and "Announcing Gremlin Foresight AI".
#DevOps #ChaosEngineering #Resilience https://opsmtrs.com/2P788Pg
@jmm86.bsky.socialFor instance, a company saw a 30% drop in server load during a non-production chaos experiment. Start small, like testing just one server. #DevOps #ChaosEngineering
@jmm86.bsky.socialA chaotic experiment too small often misses issues, like a tech team testing software fixes only in perfect conditions. It needs a wider range to spot problems. #ChaosEngineering #FailFast
@jmm86.bsky.socialKey takeaway: Observability during chaos experiments is not optional. You need to see the system behavior in real time to interpret what the failure is revealing. #ChaosEngineering
@jmm86.bsky.socialKey takeaway: Failure injection without observability is dangerous. You need to see the effect of the failure, not just cause it. #ChaosEngineering
@jmm86.bsky.socialChaos engineering shows real outages are often about latency and packet loss, not hard crashes. Try it: disrupt a single internet router. Consequences: understanding how apps respond to real network issues is crucial. #ChaosEngineering #NetworkResilience
@jmm86.bsky.socialFirst chaos experiment often blows up wider than planned. Start testing in non-production with small scope. Example: bank system rollback. Consequence: missteps delay live rollout. #DevOps #ChaosEngineering
@jmm86.bsky.socialKey insight: Mastery of blast radius is the most critical chaos engineering skill. Partitioning our network eastward recently took out half of Asia. Understanding impacts is vital. #ChaosEngineering #NetworkResilience
@infoq.comWhat happens when #ChaosEngineering meets payment systems?
Standard chaos assumes clean stops, predictable blast radius & free production testing. Payments violate all 3!
š Salim Adedeji breaks down ECS failure modes: buff.ly/JbZTkcf
@opsmatters.comThe latest update for #GremlinInc includes "How to verify your #Azure Application Gateway is zone-redundant" and "Managing #Kubernetes node drains with Pod Disruption Budgets".
#DevOps #ChaosEngineering #Resilience https://opsmtrs.com/2P788Pg
@jmm86.bsky.socialRunning chaos engineering in production taught me: The steady state hypothesis is what makes chaos engineering science instead of vandalism. Without knowing what 'normal' is, experiments prove nothing. #ChaosEngineering
@jmm86.bsky.socialAfter years working with chaos engineering ā Automated chaos in CI/CD pipelines catches regressions in resilience. Manual chaos days find problems once; automated runs find regressions continuously. #ChaosEngineering