Four ways to boost IT performance with application-aware network performance monitoring
In an era of unified IT, you can no longer afford to take a silo-based approach to monitoring and troubleshooting IT problems. It’s time for network engineers, server admins and application engineers to expand beyond their particular domains and department-specific tools. It’s time to embrace a new, integrated approach to network and application monitoring that lets you view your entire IT infrastructure from a single console and resolve issues before they affect end users. It’s time for applications-aware network performance monitoring (AANPM).
AANPM tools are network- and application-level data collectors, monitoring network devices as well as monitoring business-critical applications to establish cross-platform visibility. With AANPM tools, engineers can make better decisions while monitoring the applications and networks that, in turn, help maintain high performance of critical business applications. AANPM, though, must be used judiciously. Otherwise, you can easily drown in meaningless data and miss key factors.
So what’s the best way to use AANPM to drive down your mean time to repair (MTTR), the metric most organizations use to measure the performance of IT service providers? By measuring the following four IT performance metrics:
- Bandwidth utilization rate
- End-user page response time
- Network latency or round trip time (RTT)
- Volume of transactions processed
Bandwidth utilization rate
This metric measures the total amount of traffic for a given period of time. Typically, bandwidth utilization is measured by analyzing packet flows, and it can be tracked for an entire organization, business unit or data center. In the most basic sense, this metric expresses link utilization as a percentage of network traffic. It’s measured by using SNMP polling and flow packets (NetFlow, JFlow, SFlow, etc.).
The problems related to bandwidth utilization rates often occur when the personal, user-related activities consume excessive bandwidth, leaving very little bandwidth for business-critical applications. This can significantly lower the performance of business-critical applications and may even lead to network outages.
Benefits: AANPM tools provide engineers with capabilities such as real-time network visibility, bandwidth monitoring and traffic shaping. By using AANPM tools that collect data on both the network and the application, organizations can gain real-time visibility into where and when the network is busy. The real-time visibility helps network engineers monitor the network and prioritize which applications need additional bandwidth. These tools also help in shaping the network traffic at the network interface level with complete granularity. Therefore, by using AANPM tools, network engineers can enhance the bandwidth utilization rates for business-critical applications.
Data based on bandwidth and application usage patterns also helps managers plan and control their IT budgets.
End-user page response time
This metric identifies the time taken for the client system to process information related to the original page request. This is done by placing probes near the client system. These probes monitor turnaround time and validate if the page request has been processed in a timely manner.
Benefits: AANPM tools help you to evaluate the likely experience of users from multiple locations by identifying potential bottlenecks in resources. This way, the end-user experience of business-critical network services such as DNS, LDAP, DHCP and Mail servers can be monitored easily. Using the tools’ data, engineers can reconstruct events, analyze flow forensics for identifying the traffic on key links, and replay VoIP calls. These statistics can be particularly useful for analyzing and solving historic problems related to application performance.
Using the end-user page response time statistics collected by these AANPM tools, system engineers can also monitor and track the SLAs of service providers. These same patterns of response times can further help system engineers plan and counter any overall application outages that may occur in the future.
Network latency or round trip time (RTT)
Network latency refers to the time elapsed between transferring a packet of data from the host system to the destination system, or vice-versa. Typically, it’s measured using the metric round trip time (RTT), which refers to the amount of time taken for a packet to reach from source to destination and back again. ICMP ping and Cisco’s IPSLA would come in handy for RTT calculations.
Ideally, the RTT should be as close to zero as possible. Excessive network latency creates bottlenecks, therefore reducing the bandwidth needed for critical applications. This network latency metric can majorly impact end-user experiences. The impact of network latency on bandwidth can be classified as intermittent (lasting a few seconds) or constant, depending on the source of the delays.
Benefits: AANPM provides performance data from both network and application perspectives that includes application response time analysis and SNMP, ICMP or CLI polling data. Network engineers can assess this data to identify and resolve problem areas, which will drastically reduce the RTT.
AANPM tools help network engineers establish a baseline for network flows under three categories: stable, degrading or unacceptable flow rate. These details can also support capacity planning, as they help network engineers determine where the bottlenecks occur and which applications require more bandwidth.
Volume of transactions processed
This metric refers to the total amount of Web transactions processed during a specific period of time. If the volume of transactions is too high, they can sit in the queue for too long. This causes client systems to reprocess the transaction requests, which may result in application outages.
Benefits: AANPM helps system engineers gain an application-centric view of events happening across the network by outlining the inter-dependencies between an application and the network. It also enables engineers to identify Web transaction data by providing insights into average end-user response times, throughput, and APDEX (application performance index) scores so that the most critical paths can be prioritized over less critical ones. System engineers can then monitor and optimize the end-user experience by assessing applications in terms of how they are deployed and how they perform. This technique can directly improve network uptime and the availability of critical business applications.
Converged applications and network monitoring tools greatly enhance cross-team communication and collaboration. AANPM tools provide a single performance management interface that offers a holistic view of both network and application performance. Using this interface, network engineers gain deeper knowledge they need to tune their organizations’ networks, servers and applications. Ultimately, that knowledge can help IT departments effectively plan and monitor network traffic, server loads, and transaction volumes as well as dramatically reduce MTTR.