Requires Free Membership to View
When you register, you'll receive targeted emails designed to keep you informed of the most relevant information on Agile development, application security, testing & QA, software requirements, and more.
Hannah Smalltree, Editorial Director
A good performance monitoring solution takes tons of performance data generated by the
production infrastructure and users and analyzes it in the context of the application being
delivered. However, performance monitoring for Web-based applications is different from traditional
applications built with monolithic architectures. Web-based application architectures are tiered
with each tier residing on its own infrastructure (Web server, application server, database,
etc.)
The Web application transactions also often flow through infrastructure such as firewalls or
application accelerators, which rarely appear on the software architectural diagrams. Therefore,
the solution must be able to collect and consolidate key performance indicators from a wide variety
of infrastructure.
The solution must also provide a good understanding of the user's experience with the application,
as many Web applications are customer- and partner-facing and poor performance usually has an
immediate financial impact to the business. There are several alternatives, each with
tradeoffs.
Running synthetic transactions can quickly provide a performance baseline for a specific
application transaction, but that may not be useful for websites with complex highly interactive
applications. A correctly placed monitoring appliance can provide a good approximation of the user
experience when the application processing is conducted primarily on corporate infrastructure.
However, Web 2.0 technology such as Ajax that allows more client-side processing may necessitate
browser-based monitoring capabilities.
Another consideration is that in many cases the application user is another application.
Application-to-application transactions often occur at much faster rates, which will put enormous
pressure on the performance monitoring solution to rapidly collect and analyze data.
The solution must be able to analyze the user and infrastructure data in context. To do this, it
must understand how different types of infrastructure interact to deliver the Web application. When
the infrastructure performance data is analyzed with this relationship map, it helps IT staff to
quickly understand how infrastructure issues will affect the behavior of the application as a
whole. When the user experience data is analyzed in this context, it helps IT staff troubleshoot
more complex root causes such as architectural issues or functional errors.
This was first published in August 2008