OpenTelemetry Project Releases New Guide
What happened
The OpenTelemetry project has published a comprehensive, vendor-neutral guide to broaden the adoption of observability. The move coincides with a shift discussed on the APIs You Won't Hate podcast, where teams are increasingly using machine learning models to detect anomalies in API traffic, moving beyond static, pre-defined alert thresholds.
Why it matters
- OpenTelemetry was formed in 2019 as a merger of two competing open-source projects, OpenTracing and OpenCensus, and is now the second most active Cloud Native Computing Foundation (CNCF) project after Kubernetes. - The project's vendor-neutral approach allows platform teams to switch or use multiple observability backends (e.g., sending traces to one vendor and metrics to another) without changing the application's code-level instrumentation. - Organizations often spend 10-25% of their total infrastructure budget on observability, with logs accounting for over 50% of that cost. A vendor-neutral standard like OpenTelemetry can reduce the total cost of ownership for observability by an average of 38%. - The top corporate contributors to the OpenTelemetry project are Splunk, Microsoft, and Lightstep, with over 1,100 companies contributing in total as of late 2023. - Machine learning models integrated with OpenTelemetry data can predict system anomalies with an average lead time of 18.5 minutes, shifting teams from reactive to proactive issue management. - The OpenTelemetry Collector is a key architectural component that acts as a vendor-agnostic service to receive, process, and export telemetry data, reducing the need for multiple agents and simplifying data routing for platform teams. - A recent CNCF survey found that OpenTelemetry has a 49% adoption rate among respondents for projects that have not yet reached "graduated" status, indicating rapid growth and significant traction within the cloud-native community. - Beyond anomaly detection, AI/ML integration with observability data is being used for intelligent data retention and instrumentation optimization, with some organizations documenting a 47.8% reduction in storage costs.
Key numbers
- - OpenTelemetry was formed in 2019 as a merger of two competing open-source projects, OpenTracing and OpenCensus, and is now the second most active Cloud Native Computing Foundation (CNCF) project after Kubernetes.
- Organizations often spend 10-25% of their total infrastructure budget on observability, with logs accounting for over 50% of that cost.
- A vendor-neutral standard like OpenTelemetry can reduce the total cost of ownership for observability by an average of 38%.
- The top corporate contributors to the OpenTelemetry project are Splunk, Microsoft, and Lightstep, with over 1,100 companies contributing in total as of late 2023.
Quick answers
What happened in OpenTelemetry Project Releases New Guide?
The OpenTelemetry project has published a comprehensive, vendor-neutral guide to broaden the adoption of observability. The move coincides with a shift discussed on the APIs You Won't Hate podcast, where teams are increasingly using machine learning models to detect anomalies in API traffic, moving beyond static, pre-defined alert thresholds.
Why does OpenTelemetry Project Releases New Guide matter?
OpenTelemetry was formed in 2019 as a merger of two competing open-source projects, OpenTracing and OpenCensus, and is now the second most active Cloud Native Computing Foundation (CNCF) project after Kubernetes. The project's vendor-neutral approach allows platform teams to switch or use multiple observability backends (e.g., sending traces to one vendor and metrics to another) without changing the application's code-level instrumentation. Organizations often spend 10-25% of their total infrastructure budget on observability, with logs accounting for over 50% of that cost. A vendor-neutral standard like OpenTelemetry can reduce the total cost of ownership for observability by an average of 38%. The top corporate contributors to the OpenTelemetry project are Splunk, Microsoft, and Lightstep, with over 1,100 companies contributing in total as of late 2023. Machine learning models integrated with OpenTelemetry data can predict system anomalies with an average lead time of 18.5 minutes, shifting teams from reactive to proactive issue management. The OpenTelemetry Collector is a key architectural component that acts as a vendor-agnostic service to receive, process, and export telemetry data, reducing the need for multiple agents and simplifying data routing for platform teams. A recent CNCF survey found that OpenTelemetry has a 49% adoption rate among respondents for projects that have not yet reached "graduated" status, indicating rapid growth and significant traction within the cloud-native community. Beyond anomaly detection, AI/ML integration with observability data is being used for intelligent data retention and instrumentation optimization, with some organizations documenting a 47.8% reduction in storage costs.
