July 2020 Summaries

Kimbre Lancaster, the Director of Global Events and Field Marketing at Gremlin, has played a pivotal role in organizing significant events such as Chaos Conf and Failover Conf, the latter being a virtual conference focused on resilience during the COVID-19 pandemic. With a background that transitions from theatre acting to project management, Kimbre found her passion in creating event experiences, which led her to work in tech companies like Scalyr and Split. She highlights the similarities between project and event management, emphasizing the strategic differences between B2C and B2B events, especially regarding long-term goals and measurable success. Amid the pandemic, Kimbre stresses the importance of adaptive skills in event marketing, sharing insights on how virtual events can still foster brand connections and urging event professionals to seek roles in companies that value their expertise. Furthermore, she encourages those in the field to use their skills to support social causes, enhancing their resumes while contributing to meaningful work.

Chaos Conf 2020, the third annual conference dedicated to Chaos Engineering, is set to be held online from October 6-8 due to the COVID-19 pandemic, emphasizing the theme "Prepare for moments that matter." This year's event aims to be the largest of its kind, fostering a community around resilience and Site Reliability Engineering (SRE) best practices. The conference will feature a range of speakers from both large companies like Amazon and Google and innovative startups, with keynotes from experts such as Adrian Cockroft of AWS and Gene Kim, co-author of "Accelerate." The three-day event will include twenty sessions and hands-on workshops, focusing on reliability through practice, completing the DevOps feedback loop, and building a data-driven culture of reliability. The free online event encourages participation through RSVPs and submissions for talks, with incentives like a Chaos Conf swag pack for the first 1,000 registrants.

The Gremlin User Newsletter highlights the complexities of modern systems and the tools developed to manage them, emphasizing the role of Chaos Engineering in testing applications and the tools used to build them. AWS has introduced App2Container, a tool aimed at transforming applications into containers with deployment manifest generation for ECS and EKS, though it is currently limited to Java and .NET applications. The newsletter discusses the importance of testing containerized applications to ensure they function correctly and benefit from container environments, including autoscaling features in ECS and EKS. It also covers updates to the AWS Well-Architected Framework with a focus on Chaos Engineering and improvements in Gremlin's platform, such as the new Windows agent, updated Health Checks, and enhanced Scenario editing and targeting features for Kubernetes. By employing Chaos Engineering and the latest tools, organizations can proactively identify and mitigate risks, thus ensuring the reliability and robustness of their systems before issues affect users.

Financial institutions are transitioning from legacy monolithic systems to modern architectures like microservices and cloud computing to address challenges such as reducing latency, releasing new features, and maintaining high availability. This modernization process is complex, requiring extensive testing to avoid potential issues such as bugs, unexpected application behavior, and poor performance. Chaos Engineering emerges as a solution to these challenges by intentionally introducing failures to test system resilience, uncovering hidden failure modes, and validating environment configurations. Through tools like Gremlin, organizations can conduct chaos experiments to ensure systems withstand conditions like network latency and hardware failures, ultimately leading to more reliable and scalable applications. Continuous Chaos, involving regular chaos experiments, further helps maintain reliability by aligning with regulatory compliance and service level agreements, while also preparing teams to respond effectively to reliability issues.

Fulfillment pipelines in e-commerce have intricate dependencies, where any component failure can lead to catastrophic consequences. To address this, some architects use messaging solutions like Apache Kafka to decouple components, creating a resilient system by allowing components to publish and subscribe to message streams independently, thereby reducing the impact of service dependencies on each other. This approach, validated through Chaos Engineering experiments using tools like Gremlin, demonstrates resilience against latency and packet loss but reveals vulnerabilities when the Kafka message stream is shut down, highlighting a single-point-of-failure. The experiments show that while the Kafka-based architecture is more robust than traditional RESTful endpoints, improvements such as better Kafka cluster configurations and caching mechanisms are necessary to enhance system reliability.

Chaos Engineering is a methodology employed by companies to enhance the resilience of their systems by intentionally introducing controlled disruptions to observe how systems respond, with the goal of preventing costly outages. The blog post narrates a scenario where a retail website's checkout process failed due to a payment processor's deployment issue, resulting in significant revenue loss. Despite implementing a fix, a similar failure occurred later, highlighting the inadequacy of the solution. By using Chaos Engineering, these outages could have been preemptively identified and mitigated through experiments that simulate real-world conditions. The process involves identifying critical system paths, designing experiments to test these under duress, and observing system behavior to strengthen infrastructure. The post emphasizes the utility of Chaos Engineering in avoiding incidents that lead to downtime and financial loss, advocating for its integration into system design and testing.

Financial institutions are increasingly adopting Chaos Engineering to enhance the reliability and modernization of their systems amidst evolving financial regulations and technological advancements. This approach involves deliberately injecting faults into systems to observe and improve their resilience, which is particularly crucial in the financial sector where outages can prevent critical transactions and trades. The technique is employed to ensure smooth migration to modern architectures like cloud environments, while maintaining uptime and managing costs effectively. Chaos Engineering also supports the decoupling of complex microservices architectures, which allows for independent failure and rapid innovation without impacting customer experiences. Moreover, it aids in meeting stringent compliance requirements by rigorously testing system redundancies and updating incident management strategies. By leveraging platforms like Gremlin, financial institutions can run experiments safely, identifying and addressing availability risks before they affect users, thus enabling a balance between innovation and reliability.