A global survey of 1,300 software engineers, developers, IT leaders and executives found half of respondents (50%) are in the process of implementing an observability platform. Another 23% of respondents said that they cannot gain end-to-end observability at all, while 26% said they believe they have mature observability practices in place. The survey, published today, was conducted by CITE Research on behalf of New Relic.
The majority (94%) of IT professionals said observability is important for their job role, while nearly three-quarters (74%) said there is still room to grow their observability practice. In fact, 81% of the 450 C-suite executives that participated in the survey said they expect to increase observability budgets in the next year, with 20% expecting those budgets to increase significantly.
Overall, the survey finds that lack of resources (38%), skills (29%), understanding of the benefits (27%) and strategy (26%) are top barriers to achieving observability.
Observability, in one form or another, has always been a core tenet of DevOps best practices. Most DevOps teams have relied on monitoring tools to observe a specific process or function. The challenge is those monitoring tools track a pre-defined set of metrics such as resource utilization. Ideally, observability combines metrics, logs and traces—a specialized form of logging—to instrument applications in a way that makes it simpler to troubleshot issues without having to rely solely on a limited set of pre-defined metrics.
Nearly three-quarters of respondents (72%) noted that IT teams today need to toggle between at least two different tools to manage IT, with 13% having to use more than ten different tools to monitor the health of their systems.
In contrast, observability platforms make it possible to correlate events across both applications and infrastructure. This capability makes it easier for analytics tools to identify anomalous behavior indicative of the root cause of an IT issue. Armed with these insights, it becomes a lot simpler for IT teams to resolve issues faster.
The New Relic survey notes that 60% of respondents still monitor telemetry data at the application level, which indicates that a massive amount of valuable telemetry data is largely going unused. A full 91% of IT professionals that responded to the survey noted that observability is critical at every stage of the software life cycle, especially when it comes to planning and IT operations.
Top benefits cited include faster deployments (27%), more cost-effective use of IT resources (25%), better digital experiences (23%) and improved productivity (22%). The top use cases cited involved digital transformation initiatives (42%) followed by cloud optimization (37%), DevOps processes (36%) and bringing products to market faster (36%).
Buddy Brewer, group vice president and general manager for New Relic, said observability will become a more pressing issue in IT environments, thanks to the rise of microservices-based applications running on Kubernetes. A full 88% of IT decision-makers New Relic surveyed said they are exploring Kubernetes and containers at some level right now, but only 10% said they have deployed them in a production environment. A total of 40% expect to be in production within three years.
In general, it’s become critical for IT teams to be able to query data in real-time to resolve issues that are becoming more challenging to resolve at a time when organizations have never been more dependent on IT. Sampling data in the hopes of discovering issues is no longer sufficient, noted Brewer.
The challenge, of course, is that most IT teams have been using legacy monitoring tools for so long that it’s still hard to imagine there might be a better approach. That doesn’t necessarily mean all those legacy tools will be immediately replaced, but it does mean many of them may simply fade away over time.
One way or another, most IT teams will soon be able to observe IT environments more deeply. Developers increasingly are taking advantage of open source tools like OpenTelemetry to instrument applications in a way organizations can more easily afford, regardless of whether they fully appreciate the nuances of modern observability versus traditional monitoring.