Dynatrace goes multi-cloud with serverless monitoring • The Register

Application monitoring outfit Dynatrace has extended its platform’s support for serverless architectures to cover all the major cloud platforms, providing observability for developers working with multi-cloud application deployments.

Dynatrace said it has extended existing support for monitoring apps on AWS Lambda to Microsoft’s Azure Functions, and Google Cloud Functions, plus managed Kubernetes environments, messaging queues, and cloud databases.

Serverless applications may consist of hundreds of loosely coupled services spread across multiple environments, which can make it hard to ensure observability, Dynatrace said. Extending its support across all the clouds should make it easier for developers to analyse and troubleshoot such multi-cloud applications, allowing for greater scalability.

For many developers, AWS Lambda will have been their introduction to serverless computing, or functions-as-a-service (FaaS) as it is also known. Lambda allows users to create code that runs in response to some event or trigger, and the platform itself takes care of all the underlying resources required to run the code.

In a blog posting, Dynatrace technical product manager Patrick Thurner said AWS, Azure and Google Cloud provide a wide spectrum of serverless functions, and engineers often combine best-of-breed services from multiple sources in order to create a single application.

“It’s becoming increasingly difficult to get end-to-end visibility and real-time insights into these heavily distributed, complex environments. With the increase of interconnected functions and other services, end-to-end traceability becomes essential,” he claimed.

According to Dynatrace, one of the difficulties with serverless is limited ability to run third-party tools and limited access to the underlying infrastructure. For this reason, open observability standards such as OpenTelemetry are key to overcoming the hurdles of instrumentation.

Dynatrace builds on OpenTelemetry with an added language-specific Dynatrace exporter to enable distributed tracing capabilities such as automatic service-detection and analysis, and provides FaaS libraries that allow tracing to be added using a single line of code in functions.

All of the enhancements in the Dynatrace platform will be generally available to customers within the next 90 days, the firm said. ®