dynaTrace Continuously Monitors ShowSlow URLs

At Velocity we announced Dynatrace Ajax Edition 2.0 beta 1 (watch the Lightning demo). Besides other great features (see the complete walkthrough video) we have added support for automatically uploading measurement results to ShowSlow.com.

We have worked over the last couple of weeks to automate this process and now continuously upload performance metrics to ShowSlow. So in addition to YSlow and PageSpeed metrics, you can now also find metrics collected by Dynatrace. The data is updated every six hours and you can browse it on the ShowSlow website.

The automatic upload already uses a new feature which allows uploading results automatically to an arbitrary URL which can process the Dynatrace ShowSlow beacon. This will enable dynaTrace Ajax Edition users to set up a continuous performance management environment based on Dynatrace Ajax Edition and ShowSlow. For sure you can process this data with any other data sink as well, but ShowSlow is a great tool so why not just use it ;- ).

How does this work?

For those who are interested in how this works technically let me explain the architecture behind our automatic upload a bit. We have a Java Quartz job which continuously runs Selenium tests every six hours. We dynamically download the URLs from the public ShowSlow instance to ensure we always test all registered URLs. The browser instances are started with the Dynatrace agent enabled. Read how to do this here. After the test has finished the results are automatically uploaded to the configured ShowSlow instance. As you can see automating performance testing with Dynatrace is reduced pretty much to writing your test cases.

Automation with Dynatrace Ajax Edition
Automation with Dynatrace Ajax Edition

What can you use this automation for?

The most obvious use case is to regularly test your web applications as part of a Continuous Integration environment. However you can use it as well to monitor your public web site or even your competition ;-). I am convinced there are a number of use cases for which this makes sense and we are already looking forward to hearing what you use it for.

Wait, weren’t you talking about a commercial version?

Yes, that’s true. Later this year we will release a commercial version which is specifically targeted for the needs of more sophisticated performance management requirements. Over the last few months we got a lot of feedback from the community about what else they would need from a performance testing perspective. We have taken all this feedback and shaped a whole solution around it. Some of those requirements are pretty complex to support as you can imagine ;- ). Many of these could not be built with Ajax Edition, which was designed to be a performance optimization tool rather than an integrated solution. Fortunately we can build on the base of our commercial solution and provide answers to requirements like end-to-end tracing, automatic regression analysis and integration into big Continuous Integration environments. We will reveal more details soon ;-). At the moment we’re starting to work with a number of early access users who will be the first to try it out.

What’s next?

The automatic upload feature to arbitrary beacon endpoints will be – amongst other cool features – part of the next release of Dynatrace Ajax Edition which is scheduled for September. Why not right now? As I said there are a number of other features we want to bring into this release. As building a release is quite some work we won’t do it before all of the planned features are ready to be released. But you want it right now and cannot wait? Ok, here is the deal. If you – the community – manage to create 9857 (what a nice prime number) twitter messages with #dynaTraceAEisGreat we will release the automatic upload feature as a special release. Maybe also encourage download of the beta?] (only one message per person, RTs allowed).

Besides working on the next release we have other plans for the future as well. With our automatic uploads running, we’ve started to collect a lot of performance data. We are currently exchanging ideas how to use that data to provide additional valuable information to the web performance community. We already have a number of great ideas and our fingers are getting ready to code …

Any other questions?

As always we are happy to answer any questions and share ideas. Just get in touch with us via our forums