In the old days, the most senior members of an engineering team were the best debuggers. They had built up such an extensive knowledge about their systems that they instinctively knew the right questions to ask and the right places to look. They even wrote detailed runbooks in an attempt to identify and solve every possible issue and possible permutation of an issue.
IT teams have been relying on observability tools to (theoretically) provide intelligence and insights into operating conditions within an organization’s digital infrastructure for years. But most of these tools have come with significant shortcomings that leave IT teams wanting more.
“Overwhelming.” It was the only word Grafana Labs CEO and Co-founder Raj Dutt could use to describe how it felt to look out at the sea of more than 600 Grafanistas gathered together in Whistler, British Columbia, for the first company-wide employee event in two years.
Observability is a mindset that lets you use data to answer questions about business processes. In short, collecting as much data as possible from the components of your business — including applications and key business metrics — then using an AI-powered tool to help consolidate and make sense of this huge volume of data gives you observability into your business. Having observability for your business and applications lets you make smarter decisions, faster.
Recently I heard one of our prospects talk about a competitor who was promoting their data lake and ask, how are we different than that? His question got me thinking about why a data lake alone does not provide the depth of observability you really need. The goal of observability is to help SREs, IT Ops and DevOps teams run their IT systems with close-to-zero downtime. Consolidating data from across your environment into a data lake is certainly a good step.
If you've stumbled (or purposefully landed) on this blog post, chances are you are new to—or diving deeper—into the observability space, o11y for short. Suffice it to say, you’re not in Kansas anymore. Honeycomb in a lot of ways can serve as a yellow brick road into o11y, and this article should serve as an introduction into how Honeycomb facilitates implementing o11y into applications and distributed services.
Ben Treynor Sloss, then VP of Engineering at Google, coined the term “Site Reliability Engineering” in 2003. Site Reliability Engineering, or SRE, aims to build and run scalable and highly available systems. The philosophy behind Site Reliability Engineering is that developers should treat errors as opportunities to learn and improve. SRE teams constantly experiment and try new things to enhance their support systems.
Distributed systems open us up to myriad complexities due to their microservices architecture. There are always little problems that arise in the system. Therefore, engineering teams must be able to determine how to prioritize the challenges. Viewing logs and metrics of such systems enables engineers to know the shared state of the system components, thereby informing the decision-making on what challenge needs to be solved most immediately.
If you’re new to the concept or just trying to keep up with the conversation, Gartner defines Observability as the evolution of monitoring into a process that offers insight into digital business applications, speeds innovation and enhances customer experience. Some folks think that Observability is a new buzzword, but in fact the term was coined in 1960 by Rudolf E. Kalman, a Hungarian-American engineer.
