Google builds out automation capabilities of Cloud Trace, plans for expanded support of Compute Engine and Container Engine
Google Cloud Trace, which collects latency data from App Engine applications and displays it in near real-time for Google Cloud Platform users, received an update this week that should make it easier to track application performance and work to improve it.
The tool launched last year, and according to Sharat Shroff, product manager for Google Cloud Platform, it's currently in use by thousands of developers. The new and updated features added this week were developed based on beta user feedback.
Here's what's new:
There's more automation in the distributed tracing system. As of the update, Cloud Trace automatically instruments apps, continuously monitoring and evaluating all App Engine requests. At times, it also analyzes each endpoint-level trace in search of bottlenecks and to provide developers with additional insight into app performance.
Using generated analysis reports based on traces of the most frequently used endpoints, Cloud Trace can now determine changes in an app's latency. The tool uses latency shift detection algorithms to seek out both minor and major changes in latencies when there's a noticeable change.
The Cloud Trace API and Trace SDK can now be used to optimize the performance of customized workloads. The SDK so far only supports Java and Node.js, but a REST API has been released to support other languages.
And finally, there's an updated user interface that provides a simple and unified view of all the information Cloud Trace gathers. The latest analysis report is front and center, with a summary of latencies for each application endpoint, but users can dig deeper to gain additional insights. Cloud Trace data has also been added to Logs Viewer and the App Engine dashboard so developers can get access to it while viewing logs and when viewing App Engine project health.
Google plans to update Cloud Trace again over the next few weeks. Shroff noted in his blog post that the company will roll out expanded support for both Google Compute Engine and Google Container Engine, as well as add the ability to get more detailed auto analysis and insights.
- read the Google Cloud Platform blog