Ready for VoIP: Network Management Architectures: Network Instruments
Network Instruments, LLC, headquartered in Minnetonka, Minnesota, is a developer of distributed and affordable network management, analysis, and troubleshooting solutions.
In 1994 they developed the Observer, the first Windows-based protocol analyzer, and also the first product to troubleshoot switched environments. Since that time the Observer has grown into an entire family of network management, analysis, and troubleshooting solutions, that include capabilities for protocol and packet-level analysis, network design, optimization and mapping tools, and SNMP-based enterprise management, including embedded agents and/or probes.
All of the Network Instruments products are designed utilizing its Distributed Network Analysis (NI-DNA) architecture, meaning that all solutions are based on a single code set. The architecture includes two key components: a probe which collects data regarding a local or remote network's operation, and a console, which displays the data.
Remote probes can be installed on many types of networks, including 10/100/1000 Ethernets, WANs and wireless networks, with the control and monitoring of the probes allowed from any location, whether local or remote. Since the probe and console are derived from the same code, they work together as a seamless, integrated solution. This unified code set also allows an easier transition to additional topologies, as minimal training is required as the network grows or evolves.
The console part of the system, is called the Observer. It provides network, application, and in-depth VoIP performance analysis, with call quality scoring and VoIP metrics on a per-call or aggregate level. It can also reconstruct calls for IT staff to listen to and review.
One of the big advantages of Observer is its ability to display all applications running on the network alongside VoIP traffic, since other applications can affect VoIP quality. Observer handles real-time performance issues with over 70 VoIP-specific metrics, plus over 50 event-based and threshold-based VoIP Experts. The system provides in-depth problem identification and resolution, as well as alerts to VoIP quality issues—including alarms for unacceptable jitter level, lost packets, or alterations in the QoS stream (see Figure 1).
The console is available in three versions.
The Observer Standard can capture, view, and decode over 590 different protocols of network traffic in real time, analyze the network traffic to diagnose critical problems, and collect long-term trending statistics for proactive management operations.
The Observer Expert can also track the network applications, and includes an extended level of VoIP analysis, conversation tracking, and performance predictions. For VoIP applications, it can monitor over 70 call quality metrics.
The Observer Suite provides all of the functionality of the Observer Standard and Observer Expert, plus SNMP device management, and remote monitoring (RMON) management.
Similarly, two options are available for the probe side of the architecture.
The Advanced Multi-Probe reports to Observer consoles located anywhere on the network for distributed visibility, and provides real-time decodes, statistics and long-term trending data.
In addition to all of the capabilities of the Advanced Multi-Probe, the Advanced Expert Probe displays remote expert analysis in real time for faster troubleshooting, performs packet captures and decodes at the individual probe level, and conserves bandwidth by transferring only Expert screen shots to the console, not raw data packets.
Two other Network Instruments management products are also worth mentioning.
The GigaStor provides retrospective network analysis by capturing everything traversing the network—every packet and every connection—and stores terabytes of network-level data for later review, analysis, and reconstruction. Think of this as a TiVO-like device for your network, giving you visibility into events even if you're not there and providing playback whenever you choose to review the activity. It also eliminates the guesswork from VoIP troubleshooting. Rather than having to recreate or guess at the cause of a VoIP issue, you replay the time period around the event and have the Observer run Expert Analysis on the packets to diagnose the issue (see Figure 2).
Another complement to the Observer is the Link Analyst which is focused on infrastructure, and is an active system for querying devices status across the network. Link Analyst displays by business group, location, department, device type, or other custom classification, the status and performance of the network infrastructure. At a high level, you can immediately tell whether every device responsible for providing VoIP service across your network is properly functioning, and if a device is failing, you can drill down on it for specific information or errors (see Figure 3).
Further details on the Network Instruments architecture and products can be found at http://www.networkinstruments.com/. Our next tutorial will continue our examination of vendors' network management architectures.
Article courtesy of Enterprise VoIP Planet