While at Cisco Live today, I was struck by the traffic patterns on the show floor. Wherever there was a traffic jam, it seemed as though it was caused by a few people walking slower than everyone else, or by a momentary obstruction that halted traffic. Enterprise networks share many of the same attributes (and problems) that show floors do in that respect. The part that makes it worse for enterprise networks (vs. show floors) is that there are mission-critical applications that run on top of these networks. When networks have performance issues (even momentary ones), the impact on these applications can be catastrophic.
Like traffic problems, simply finding the bottleneck doesn’t necessarily mean you found the problem that caused the bottleneck. Show attendees can slow down their pace because they are reading a booth banner or sign (good for the people with that booth, but bad for traffic), because someone tripped on a loose piece of carpet, because they are speaking on the phone, or for any number of reasons. Unless you had a video recording of the time around the bottleneck, you are forced to “guess what happened” based on what people saw or experienced. Networks have exactly this same issue, only of a greater magnitude and higher complexity. Relying on things such as SNMP traps, pings, or sampled network data to identify the root cause of an issue, is a lot like identifying the source of a traffic problem – it leaves way too much to interpretation and guesswork. Worst case, you may have to instrument your network and hope that the problem happens again (not a great situation to be in, let alone to explain to your boss).
This brings me to Cisco Live, where Emulex and Dynatrace (formerly Compuware) are jointly announcing a new solution for application-aware network performance management (AA-NPM): The Emulex Endace FusionTM Connector for Dynatrace’s Data Center Real-User Monitoring (DC RUM) product. The Endace Fusion Connector for Dynatrace’s DC-RUM provides the high-fidelity network playback that allows network operations (NetOps) personnel and application specialists to actually see what happened before, during, and after the event by providing 100 percent of the relevant network packets to the Dynatrace toolset. This data is not only available “when it happens,” but is also stored on disk in the EndaceProbeTM Network Recorder, allowing further “after-action” analysis to occur if needed. This data is available to the Dynatrace Agentless Monitoring Device (ADM) applications, which can actually run on the EndaceProbe (further reducing the footprint of the solution in the data center), as well as to Dynatrace’s Transaction Trace Analyzer (TAA). This combination provides a highly scalable Application Performance Monitoring (APM) solution for high-density 10Gb Ethernet (10GbE) networks running network-centric mission-critical applications, and significantly streamlines workload for NetOps and application analyst personnel troubleshooting critical issues.
If you want to hear how this solution can help you get out of the “guessing game” and solve your critical application performance issues the first time around, please contact us and tell us about your environment and some of the issues you have run into. Whether you are running a mission-critical database, a VDI application, or some other mission-critical application, we can help you ensure the performance of your applications. Looking forward to talking to you!