Datadog is digging into serverless computing, with what it claims will be end-to-end monitoring for AWS Lambda functions.
The New York-based company’s raison d’etre is to offer the ability to “see inside any stack, any app, at any scale, anywhere.” However, the nature of AWS Lambda, and other serverless architectures makes this difficult – you’ve ditched the “servers” and handed over responsibility for operations and scalability to the cloud provider after all.
Or as Datadog puts it, “…traditional instrumentation methods cannot be deployed. Without insight into this infrastructure, DevOps teams face blind spots where performance issues can occur for applications that rely on AWS Lambda.”
Datadog has gotten round this by integrating with the AWS X-Ray distributed tracing system, while a new Cloud Functions View in Datadog will “pull together Lambda metrics and logs with distributed request traces from your functions”.
The Cloud Functions view will allow users to search and sort functions, and view stats such as error counts, memory usage, etc. In addition, said Datadog, “AWS X-Ray traces application requests from end-to-end as they traverse your cloud infrastructure.“ This is achieved by enabling active tracing in the Lambda console, and importing the X-Ray SDK into the function code.
There’s more than one type of function of course. A spokesman for Datadog told us it supports metric and log collection from Azure Functions and Google Cloud Functions, “and we will have support for both in the Cloud Functions Page in the future.”