Applying and Maximizing Observability with Christine Yen

| | |

TCP Talks - Christine Yen

About TCPTalks

TCPTalks are a different format of The Cloud Pod podcast, where instead of discussions about the latest news and events, the hosts give the stage to an expert in a specific field of cloud computing. It’s a chance to dive deeper into the technologies that are emerging, and a great place to learn more about specific interests in the field.

This Week: Applying and Maximizing Observability with Christine Yen

The term ‘observability’ refers to how well we can observe a system’s inner workings based on its outputs. Observability can apply to any given system, from software to mechanics to engineering, but here it is mainly applied to the systems that run a business. And since business analytics essentially exist with the purpose of granting observability to decision-makers, understanding how to increase it can be a world-changer for global operations.

Christine’s software, Honeycomb, is a data store and query engine that “are purpose-built to detect patterns across billions of requests in under 3 seconds, even with highly granular data” That means Honeycomb is an observability system specifically made for software developers, and that its co-founder Christine has dedicated her professional interests to the concept. 

Christine breaks down how a company can apply observability to DevOps and code data to find out why software isn’t working as expected, or to verify that it is. Looking deeper, she also discussed how an observability platform like Honeycomb is used with AWS to extract data directly from website and application behavior. The data is used like any metric–to identify key goals for performance improvement–except Honeycomb is checking code behavior in the application layer instead of the front or back end.

As the results of observability only become apparent if engineers are asking the right questions, Honeycomb’s current endeavor is to figure out how AI can help them along. She discusses how the challenges are the variations in how engineers record and measure data on different teams, and how many are used to making their own observations directly. Finally, she expressed how cultures that are already versed in the right questions can really benefit from Honeycomb’s applied observability.