A survey of 1,255 observability practitioners and IT leaders finds that reliance on open-source software to enable observability continues to increase, with 76% of respondents now using open-source software tools and platforms such as OpenTelemetry and Prometheus.
Conducted by Grafana Labs, the survey specifically finds that 70% of respondents are using both Prometheus and OpenTelemetry in some capacity, with half working for organizations that increased their investments in both technologies for the second year in a row.
More than two-thirds of respondents (67%) said their organization uses Prometheus to monitor applications in production environments, while OpenTelemetry is being used by 41% of respondents to collect telemetry data. More than a third (38%) of respondents are investigating OpenTelemetry, with 6% reporting they have no plans to use it at all.
Grafana Labs CTO Tom Wilkie said the survey makes it clear that open-source platforms that don’t require a schema to organize the massive amounts of telemetry data being generated by modern applications are increasingly becoming foundational tools for enabling observability at scale. That approach makes it possible to analyze telemetry data without first having to curate it, he added.
In fact, as observability becomes a more mainstream IT tool, more senior leaders are involved in making platform decisions, he noted. For example, approximately three-quarters of all respondents said observability has become business-critical for the CTO, vice president, or director level, with CTO being the most common response (33%). That level of interest indicates that observability has now become a C-level concern, especially as costs continue to rise, noted Wilkie.
Overall, the types of telemetry data being collected span metrics (95%), logs (87%), traces (57% and profiles (16%), with organizations on average relying on eight different technologies to collect this data. The most important factors when evaluating observability tools and platforms are cost (75%), followed by ease of use (58%), while the biggest concerns cited are complexity (37%), signal-to-noise ratio (38%) and costs (37%).
The most sought-after new capabilities are artificial intelligence (AI) capabilities that fire off alerts based on changes to metrics (31%) and faster root cause analysis (28%). It’s now more a question of how soon, rather than if, artificial intelligence (AI) agents will be applied to address those issues, said Wilkie.
The survey also finds that well over half of respondents (57%) are working for organizations that mostly or only self-manage these tools, versus 37% that are relying on some type of software-as-a-service (SaaS) platform that is managed by a vendor on their behalf.
Each IT organization will need to determine its own best observability path forward, but the one certain thing is that the number of end users interested in this data now goes well beyond IT teams. Business leaders are starting to correlate IT events to business outcomes in ways that enable organizations to respond faster to changing business conditions.
The challenge, as a result, is finding an approach to observability that can be widely employed in a way that doesn’t break the IT budget, in an era where more organizations are sensitive to the total cost of IT than ever.
