Google Wednesday released a beta version of a software-debugging tool that it says will give application developers a way to quickly diagnose performance bottlenecks in production applications running on the company's cloud computing platform.
Google Cloud Trace works with applications running on Google App Engine or using a Google Cloud Platform managed virtual machine. It helps developers gain visibility into software code while it is running, to see which line or lines of code might be dragging down application performance. In the event of a performance bottleneck or other problems, the tool generates a report containing details on what might be causing the issues.
“With Google Cloud Trace, you can diagnose performance issues in your production application by quickly finding the traces for slow requests and viewing a detailed report of where time is spent in your application while processing these requests,” Google product manager Praful Dublish wrote in a blog post.
It helps developers find persistently slow requests that may cause sluggish application performance even if it only affects a small number of users, Dublish said. The software tool can also be used to compare performance levels between an old version of an application and a new one.
Further reading
Google announced Cloud Trace at its I/O event last June and has been making early versions available for testing on an invitation-only basis. With today’s announcement the beta version is available to anyone who is developing an application or has an application running on Google’s Cloud Platform.
Google Cloud Trace is one of three new tools that the company is working on to help developers catch and eliminate problems in operational applications. The other two are a Cloud Debugger tool that lets developers set and monitor a ‘watch point’ in their software code for performance issues and a Cloud Monitoring tool that lets developers find and fix performance problems quickly by giving them better visibility into Google’s Cloud Platform services..
The impetus for the effort comes from what Google says is the need for better tools to help developers resolve problems in production cloud applications that are already deployed and running. Unlike applications that are still in the development and testing stage, production applications are harder to debug because it often involves bringing down a system that is being actively used.
Application error logs also often do not always contain enough information to help developers isolate the root cause of a problem especially in situations where a problem might affect only a small fraction of users. So diagnosing performance problems in production application can be extremely difficult, Dublish said.
“Cloud Trace helps you visualize and understand the time spent by your application for request processing. This enables you to quickly identify and fix performance bottlenecks,” he said.
