Today, Timescale, the creators of TimescaleDB - the category-defining relational database for time-series - announced the release of support for OpenTelemetry traces in Promscale, the observability backend for metrics and traces powered by SQL (read the full blog post). Promscale ingests and stores OpenTelemetry, Jaeger, and Zipkin traces and provides integration, analysis, and visualization support for existing open-source tools, Jaeger and Grafana. Based on TimescaleDB and PostgreSQL, Promscale offers full SQL capabilities and provides users the power and flexibility to interrogate their data and answer any question with a language most developers already know. With the new support for traces, Promscale provides a single database and unified query language for metrics, traces, and business data.
Increasingly, developers build their applications as microservices running on highly dynamic Kubernetes and cloud infrastructure. Microservices enable developers to build and release software faster and shorten the feedback loop with customers. However, it leads to complex systems where tens or hundreds of components are required to process an individual request. Those systems are much more difficult to understand and operate because their behavior is more dynamic, and failure patterns are much harder to predict. Traces complement metrics and logs and provide better visibility into microservices environments by representing how an individual request flows through components in a distributed system.
With traces, it is possible to build a connected representation of all the steps to process individual requests to identify the service and operation causing issues. It’s also possible to get a holistic view of the dependencies between all the different components of the system. Unfortunately, the adoption of traces has been slow because they have traditionally required a lot of manual work to implement and a new backend to store them. In addition, existing open-source tracing solutions provide limited ways to analyze traces.
OpenTelemetry, an emerging CNCF standard for instrumentation, provides vendor-agnostic SDKs, libraries, and tools that reduce the time developers need to get their services instrumented with traces. But to realize the value out of trace data, developers still need flexible ways to query and aggregate traces (and other data) to answer any question they need to answer about the performance of their applications.
Promscale addresses these challenges and supports OpenTelemetry. The new support for traces in Promscale provides developers with a single observability backend and full SQL query language for their Prometheus metrics and OpenTelemetry traces. Promscale now eliminates the complexity of installing and operating separate backends for metrics and traces and opens unlimited possibilities to query traces. Developers can now answer questions impossible to answer with existing open-source tracing tools, such as:
- Identify the service operations with the highest error rate
- List the slowest database calls across all services
- Find the upstream service causing the load on a service that is seeing elevated load
Even more, with Promscale and SQL, developers can correlate traces and metrics at query time, allowing them to, for example, show response time per service alongside their container CPU and memory consumption.
“Understanding the behaviors and failure patterns of microservices on cloud-native infrastructure is difficult with just metrics and logs,” said Ramon Guiu, VP of Observability, Timescale. “Yet most developers are not adopting traces. First, it’s too much work because they need to instrument their services manually and set up and operate a separate backend to store them. And second, the tools they have today don’t help them get a lot of value out of that trace data. Promscale with OpenTelemetry provides developers with a solution that is much faster to implement with automatic trace instrumentation and a single storage system for metrics and traces. It also provides the benefits of a complete SQL query language to do deep analysis on their traces and correlate it with metrics and business data to help them build better software.”
Support for OpenTelemetry traces is currently in beta and available today at www.timescale.com/promscale
At Timescale, we are dedicated to serving developers worldwide, enabling them to build exceptional data-driven products that measure everything that matters: software applications, industrial equipment, financial markets, blockchain activity, consumer behavior, machine learning models, climate change, and more. Analyzing this data across the time dimension ("time-series data") enables developers to understand what is happening right now, how that is changing, and why that is changing. Timescale develops TimescaleDB, the category-defining open-source relational database for time-series data, and Promscale, the observability backend for metrics and traces powered by SQL. Timescale offers fully-managed TimescaleDB and related database services. Timescale is a remote-first company with a global workforce and is backed by Benchmark Capital, New Enterprise Associates, Redpoint Ventures, Icon Ventures, Two Sigma Ventures, and other leading investors. For more information, visit www.timescale.com or follow @TimescaleDB.