In APMdigest's exclusive interview, Doug Roberts, Director of Enterprise Products at Fluke Networks, discusses Application Aware Network Performance Management (AANPM), how it works and what benefits this solution delivers.

Start with Part One of this interview

APM: How does AANPM actually work? How is the network made aware of the app?

DR: AANPM works by starting first with an understanding of how well applications are traversing the network. That means monitoring the actual transactions end users are performing and measuring how quickly the network, server and application deliver the data to the end user. This application awareness comes from capturing the packets on the wire and monitoring their Quality of Experience (QoE) in real time.

The next level of visibility is to understand not just how quickly the transactions occur, but who the user was, and what they were trying to accomplish. This means performing Deep Packet Inspection, or DPI, to analyze what the application was, and what function the end user was performing when they experienced the issue.

This allows the IT team to start troubleshooting performance problems with an understanding of who the user is, what they were trying to do, and where the delay occurred (server, network or application). Response time data is then tied into workflows that leverage traditional NPM data to look at flow, rate, utilization, endpoint details or even raw packet captures along the path of the user or application.

APM: Who are the users of AANPM within the IT organization?

DR: Anyone with a stake in how applications are performing is a potential user of AANPM. For most organizations, the network team becomes the defacto starting point when slowdowns occur, but since AANPM provides metrics in terms of the applications that everyone supports, it’s easy for multiple teams to leverage the data. Typical users depending on organization structure include: network teams (WAN, Wireless), server teams, application owners, DevOps, monitoring/operations teams.

APM: What are the key benefits of AANPM?

DR: AANPM offers several benefits to IT teams, including the ability to: quickly isolate performance problems to avoid war room troubleshooting and finger pointing; reduce MTTR through streamlined troubleshooting across multiple data sources; reduce IT spend through intelligent capacity planning.

Understanding how quickly the network, server and application are responding to user requests allows IT to immediately identify the problem domain and start troubleshooting the root of the issue. This avoids the classic “war room” scenario so common within the industry.

Additionally, by providing supplemental data sources such as SNMP, flow, packet analysis, path analysis and synthetic testing in structured logical workflows, IT can quickly able to isolate root cause. This leads to reduced mean time to repair (MTTR).

Leveraging AANPM data also allows for more intelligent decision-making on the part of IT engineers, managers and directors. With AANPM, they can quickly identify not just where the highest utilization is, but also what comprises that utilization and what the quality of experience at the other end of that link may be.

Additionally, once upgrades or IT projects are deployed, AANPM allows IT to verify that performance has stayed within the baseline or has improved in an expected manner. Did adding more memory really speed up that server? Is traffic from that new application affecting our existing critical applications? Without AANPM, these are “gut feel” judgments.

APM: What does Fluke Networks mean by "end-to-end AANPM"?

DR: Our definition goes a little further than most. The “last mile of visibility” relates to the remote, branch office or individual user (VPN). The Fluke Networks AANPM solution provides an understanding of response time and performance all the way to an end user and his or her specific device, with full visibility into network health along the entire path taken by the end user.

APM: How is Fluke Networks AANPM differentiated in the market?

DR: TruView delivers five capabilities in one single appliance and a single user interface: Response time, traffic analysis, device performance, VoIP QoE and retrospective packet analysis.

And we provide this in an appliance, which is:

■ Simple: racked to reporting in under 15 minutes; auto-discovery and configuration; intuitive web interface.

■ Intelligent: self-learning baselines; time-correlated views; guided workflows.

■ Complete: monitoring and troubleshooting; network and application; packets, flows, SNMP, active test.

ABOUT Doug Roberts

Doug Roberts is the Director of Enterprise Products for Fluke Networks, and has been with the company since its founding. He has worked as an IT professional for over 18 years in various technical lead, business development, and product innovation roles. Currently Roberts leads the product strategy team and has been instrumental in developing the next generation of network and application performance products for Fluke Networks. Roberts is active in the IETF, IEEE PCS, and the APM Forum, with 9 patents (issued & pending) in the areas of response time measurement, big data storage and retrieval, and application efficiency measurement logic. He has provided his subject matter expertise in network troubleshooting, application performance monitoring, and transaction analysis to most major Service Providers in the world, along with thousands of private enterprise owners through guest speaking events that include Interop, Cisco Live, and TechNet. Roberts holds degrees in Engineering, Business, and Statistics from The Georgia Institute of Technology and Mercer University. In addition to his formal education he also holds a myriad of both technical and industry certifications. Prior to joining Fluke Networks, Roberts worked for Sniffer University.