Network performance monitoring is the process of visualizing, monitoring, optimizing, troubleshooting and reporting on the health and availability of your network as experienced by your users. Network Performance Monitoring (NPM) tools can utilize different types of telemetry, including:

• Device metrics, such as SNMP, WMI, CLI, API, logs, synthetic tests.
• Network flow data, such as NetFlow, jFlow, IPFIX, etc.
• Packet data. Everything you do on a network requires packets; it carries data across the network.

How does Network Performance Monitoring work?

Network Performance Monitoring solutions traditionally collected data from a variety of sources: SNMP, flow data, and packets.  Each provides a different perspective on the problem that when combined, provides a complete understanding of the health of your network and the applications running over it. Network Performance Monitoring solutions are typically available as hardware, virtual and cloud software so you have complete visibility across hybrid or multi-cloud environments.

SNMP

Simple Network Management Protocol (SNMP) is used to manage and monitor network devices and their functions. One of the most widely supported protocols, it provides the information necessary for fast detection of network infrastructure outages and failures. SNMP delivers critical information and network diagnostics about network device and interface availability and other performance indicators, such as bandwidth utilization, packet loss, latency, errors, discards, CPU, and memory. SNMP is supported on an extensive range of hardware—from routers and switches to endpoints like printers, scanners, and internet of things (IoT) devices.

Flow data

Flow data is generated by network devices, such as routers, switches, and other devices. Like a phone bill that tells you who you spoke to, how long you talked, how often, etc, Flow data provides similar information. We know who communicates with whom, when, how much data was transfer across the network, how long, how often, etc.; but we do not know what the subject of the conversation was. 

Benefits of using NetFlow for monitoring network traffic

Monitoring and analyzing flow data helps obtain valuable information about network users, peak usage times, and traffic patterns. In contrast with SNMP data, flow-based network monitoring can understand the traffic patterns, provide a holistic view for monitoring network bandwidth utilization and WAN traffic, support QoS validation and performance monitoring, and it can also be used for network troubleshooting.

Full packet capture

Packet capture is the term for passively copying a data packet that is crossing a specific point in a network and storing it for analysis. The packet is captured in real-time and stored for a period of time so that it can be analyzed. Packets help diagnose and solve network problems such as:

• Troubleshooting detrimental network and application activities
• Identifying security threats
• Troubleshooting network errors, like packet loss, retransmissions
• Understanding capacity issues
• Forensic network analysis for incident response

There is a saying at Riverbed that “Packets don’t lie.”  A packet consists of two things: the payload and a header. The payload is the actual contents – a voice call, an email message, etc. While the header contains metadata, including the packet's source and destination address. Entire packets or specific portions of a packet can be captured and retained.

The Riverbed Solution

The Riverbed Unified NPM unifies device monitoring, flow monitoring, and full packet capture and analysis solutions. These solutions are tightly integrated together so that you can more quickly troubleshoot complex performance issues. They also integrate into the Riverbed Portal that combines them into collated, easy-to-use dashboards that streamline analysis of complex problems.

By blending APM, EUEM, and NPM data from industry-leading tools, Riverbed creates a dynamic map of network and application performance. Different business and IT teams can gain a complete picture of the environment. Instead of wasting valuable time and resources, you can rapidly diagnose and fix service issues before end-users notice.