[ This article is part of a series promoting FailoverConf -- a virtual event dedicated to resilience hosted by Gremlin on April 21. Join! ]

By Jeff Lo, Director of Product Marketing at Splunk

Traditional IT teams and application developers approached building resilience into their infrastructure and applications in a very waterfall way. They would lean on documentation, force change requests, and try to anticipate all that could go wrong before testing what had been built. The complexity of modern tech environments is such that even with the most diligent of development teams, there is no predicting how systems will fail, and there is no question that things will go wrong in the most unexpected ways.

In today's fast-paced and competitive business environment, attempting to plan ahead for every possible disaster scenario is far greater than the risk of not deploying code. Those who maintain application velocity are often those who will win in markets where customer demands are high.

That is why resilience needs to not only be tightly integrated into development, and testing, but also be baked into the tools that monitor and support the applications in production. This means that monitoring tools need to see all aspects of the application from the back-end infrastructure to the front-end user. They also need to detect and surface issues in a timely and relevant way. And they need to provide context that does not force those who are on-call to hunt for details related to the issue.

Only with robust observability can teams maintain application velocity and support the crucial feedback loop that is necessary to build stability into applications and the infrastructure that runs it. For legacy applications an a priori approach to resilience may have worked. But modern application architectures are complex and consist of many small interrelated stateful and stateless services, with greater service dependencies, greater release velocity, and even more disparity between environments. All of this means operations and SRE teams need to expect failure, respond quickly and learn. In addition to a robust and automated testing strategy, choosing processes and tools that set guard rails for resilience and support feedback loops is necessary to modern environments.

Splunk supports data-driven application development with a suite of observability tools that increases visibility from planning to production. For modern, cloud-native applications SignalFx Infrastructure Monitoring and SignalFx Microservices APM solutions are able to keep pace with modern architectures built on containers, Kubernetes and microservices, and deliver insights that improve the performance and resiliency of all services in those applications. With these solutions, Splunk gives today's DevOps teams the confidence to deploy code quickly and reliably.

Additionally, Splunk's incident response solution from VictorOps connects incidents across modern and legacy systems, and helps mobilize the right on-call engineers while providing them the context they need to address issues with speed and precision. Splunk Enterprise gives organizations high-fidelity data to support troubleshooting of the most complex technical issues, and Splunk gives DevOps teams the ability to tie delivery chain metrics to business value. All of Splunk's technologies are backed by AI and Machine Learning to reduce the human time needed to respond to issues, detect issues even before they happen, and help spot abnormal behavior in real-time. Join us at the virtual FailoverConf on April 21st to learn more about resilience from other industry leaders.

##

About the Author

Jeff Lo is Director of Product Marketing at Splunk with 20 years of experience in product marketing, product management and go-to-market. Prior to SignalFx, Jeff led Product Marketing at Scalyr and ran global product marketing for Predix Studio and Digital Twins at GE Digital. Jeff holds a B.Sc. in Electrical Engineering from the University of Alberta.