The “Argument Ender”: SteelCentral Transaction Analyzer Tracks Down Performance Bottlenecks


If you run a complex network infrastructure, you know that troubleshooting application performance problems can give rise to finger-pointing between different silos on your team: Is the problem in the network, the database, or the application server?

The good news: Riverbed SteelCentral™ Transaction Analyzer (known as the "argument ender") quickly delivers the data you need to track down performance bottlenecks in multi-tier application transactions that span your network and application infrastructure. Among other benefits, with SteelCentral Transaction Analyzer, you can:

  • Automatically diagnose network and application performance problems
  • Conduct "what if" analyses to predict the performance impact of network or application changes

And now it's been updated to meet modern security requirements and work with more operating systems. Let’s examine typical use cases in detail, and then take a look at how Transaction Analyzer’s lightweight agents work and its tight integration with SteelCentral AppResponse 11.

Ending arguments

Transaction Analyzer ends arguments amongst your staff by answering their most pressing questions about performance delays. Is the problem in the database? Is it the client? Is it in the application itself? If so, what tier of the application is it in? Is it in the network? If so, what type of network problem are you dealing with?

Of course, often you're dealing with more than one problem at once. The pie chart in Figure 1, an example of the sort of information Transaction Analyzer can deliver, quantifies the different aspects of the application transaction and the network that contribute to application response time.

Figure 1. Transaction response time, segmented

Testing those apps before you unleash them

Many Transaction Analyzer customers also use the tool to test applications in a controlled environment before they're rolled out to users. Its ability to quantify performance is great for development teams who need to meet specific SLAs.

Results can be used by application teams to give an indication of how well the application can be expected to behave in their existing network infrastructure, and by network teams so they can plan for what’s coming. And deployment teams can analyze application transactions on various network infrastructures to see how that affects performance.

Figure 2 shows sample results of this sort of testing, with application response times broken down across network and application tiers for various deployment scenarios. The dashed line down the middle represents the SLA the developers are testing against. You can see that in this example, Transaction Analyzer proves that moving the application to the cloud will degrade response time below acceptable levels but that adding WAN optimization will fix that problem whereas just upgrading to a faster WAN link won’t.

Figure 2. QuickPredict used for predicting application performance

Tracking down trouble

With Transaction Analyzer, moreover, you can create formatted, ready-to-share reports with only a few clicks that summarize the analysis for easy consumption by various application stakeholders. One of the main purposes for such analysis is to find the locations and causes of bottlenecks and performance problems in your application infrastructure. For instance, the analysis could help identify an individual SQL statement or network protocol that's slowing down your application.

Your agents on the job

SteelCentral Transaction Analyzer Agents (also called capture agents) are lightweight software agents that can be installed on existing hardware (servers, desktops, VMs, cloud instances, etc.) and are used to collect packet traces in a ubiquitous and highly distributed manner. These packet traces are imported into SteelCentral Transaction Analyzer to create a transaction model for deep-dive analysis, prediction, and troubleshooting.

Works hand in hand with AppResponse 11

To import packet traces into Transaction Analyzer, you can take advantage of the tool's close integration with AppResponse 11. If you find an anomaly in the AppResponse UI, you can right-click and quickly and seamlessly drill down from a specific insight or view straight into Transaction Analyzer. From there, you'll be able to take a deep dive right to the packet level — an ability that differentiates SteelCentral from anything else available on the market.

Fully up to date and ready to go

SteelCentral Transaction Analyzer has recently been updated to make it even more secure, and to build compatibility into Transaction Analyzer Agents for multiple operating systems, including Windows 10, Enterprise Linux, and macOS.

Transaction Analyzer is available in two models: SteelCentral Transaction Analyzer and SteelCentral Transaction Analyzer Plus, which includes Packet Analyzer Plus, a holistic packet analysis tool that offers many views into a given packet capture. For more information about both models, visit the product website.