BOSTON, MA – October 01, 2012 – Spark::red started offering New Relic™ as a part of its Spark::red™ Oracle ATG Commerce Solution. New Relic web monitoring tool provides an innovative solution, merging end user performance and satisfaction monitoring, tied to back end application profiling and performance analysis. An addition of such a tool allows Spark::red’s eCommerce customers to increase their web performance as well as to improve customer satisfaction.

The end user monitoring tracks page load times for the website users, analyzing time spent in the web application, the network, the browser DOM parsing, and the browser page rendering using a combination of synthetic clients and tracking all incoming page requests from real users. User happiness is calculated based on page render times with configurable thresholds. You can also set up alerts for performance threshold violations.

You can see historical data and trends for your site or for a specific page or URL.

This makes it easy to see what impact various events have had on end user page load time and satisfaction: new EAR deployment, infrastructure changes, high traffic spikes, and more.

Where New Relic™ really differentiates itself from other user satisfaction performance monitoring services such as Keynote or Gomez, is the server-side plugin that runs in your JBoss application server. This plugin allows you to see application server performance data as a whole, by JSP page, or by specific page request. This helps you to see an alert or slow trend on the end user side of the house, and click straight through into the full call tree of the slow page/requests. You can see where time was spent: slow SQL, JSP includes, servlets, and more. This lets you easily pinpoint the actual performance bottleneck for a single request event, or examine trends of commonly slow interactions. You can also enable full JVM profiling, which typically comes with a massive performance penalty for the entire time a JVM is running, on the fly, as needed.

In short, New Relic™ allows you to instantly know if you have slow pages or interactions on your site, and then quickly discover precisely what JSP, Java, or SQL is contributing to the slowness. Better still, it boasts an integration to Jira, allowing you to instantly create a Jira work ticket for your dev team to remediate any bottlenecks you discover.

As an example:

You might see that end users are seeing a specific product page as being slower than you’d like. You can quickly dive in and see that the slowness is from the application layer, not the network or browser parsing. You can dive back into the application server and see aggregate average data for that page, or view a single specific page transaction. In either approach you can see what specifically is taking the most time, in this case perhaps the particular product has a very large number of SKUs and your application is doing real time inventory look-ups to the database for each SKU. You can see the code path through JSP, any JSP fragments, and the code behind it, down to the database calls and the specific SQL queries being generated. This might be a great place to introduce distributed hybrid repository caching for your inventory repository, or just add a Cache droplet with a 5 minute timeout, if you don’t need fully real-time data.

As a result, through utilizing New Relic you can increase your website performance as well as ensuring happy and loyal customers.

Black Friday & Cyber Monday-2012 Online Shopping Behavior Forecast  2

In our real world use of New Relic with several production ATG websites, we’ve not only been able to quickly identify low hanging fruit with massive performance gains, but also been able to track down root causes to corner case issues that don’t exist in the test environments. It’s a tool that pays for itself very quickly, and is the very best in performance monitoring services.