VMware Tanzu recently released The State of Observability which is a survey that collects information on how developers view observability.

This year survey collected data from 315 working professional, all of who are responsible from mission-critical cloud applications in a DevOps, SRE, Architect, Application Delivery, or Tools roles in a company with 100 or employes with at least 10 developers. All of the organizations that surveyed have a significat software development footprint.

Here are a couple key takeaways from the report.

Observability momentum

• Observability has been gaining quite a lot of momentum in the past year as more and more developers are already using, implementing, or in the process of evaluating various tools.
• Many developers believe that implementing observability will benefit their organizations.
• Among devlopers who have been using observability tools, many say that it is a absolute must for modern cloud applications.

Growing cloud-native application complexity

• The adoption of observability is increasing as enterprise IT operations are using hybrid and multi-cloud applications.
• A lot of the respondants beleive that modern cloud applications are significantly more complex.
• Developers are rapidly releasing or making updates to critical cloud applications several times a week.
• Teams are moving to a microservice architecture and increasing their delivery speed, all of which demands the need for observability.

Perception shift of observability

• Nearly all of the surveyees believe that observability capabilities would benefit their organizations due to features like easy integration, multi-source monitoring, and syntentic and real user monitoring.
• Devlopers agree that cloud services and applications would have better availability and performance if more of the team had visibility into the infrastructure and application metrics.
• The top challenges with visibility are lack of unified visibility across all services, limited insight into cloud resource usage and limited access to teams.

Challenges with tools

• There is no way to visualize and co-relate application and infrastructure metrics which makes troubleshooting difficult.
• Some desired monitoring features are monitoring multi-sourced data at scale, metrics such as histograms, traces and span logs, and co-relating customer and bussiness metrics to application and infrastructure performance.
• Distributed tracing ranked as the most beneficial observability capability.
• There are way too many tools which offer different functionalities and a lack of a unified toolset.