Even the most modest network has complex monitoring needs. Administrators must monitor the overall performance of the network, the status of individual devices, application and server performance, traffic and processor loads, and even the physical environment (such as server-room temperature, flood risks and so on). Network OSes can log statistics and events, but manually reviewing these voluminous logs is overwhelmingly tedious. Fortunately, there are many commercial and open-source network-monitoring tools available to automate the process. Before you buy any network-monitoring tool, however, you need to define what your company's needs are and what features you desire.
Take an inventory of your company's network to identify the devices and applications that you need to monitor. This will help you define the things that a network-monitoring solution should observe. Decide what network events will require alerts to be sent to administrators, who will receive each type of alert and how you want the alerts delivered (via email, a text message to a cell phone and so on.). Once you have determined these factors, you will be able to zero in on the right network-monitoring tool.
Some of the metrics that a solution should monitor include RTT (round-trip time), the time it takes for packets to travel from point A to point B and return (including response processing time at point B); application-response time (which varies depending on traffic load); packet loss; "jitter" or variations in RTT; unauthorized access attempts; processor load; operating temperatures of processors and other devices; queuing delay; and so forth.
Many network-monitoring tools have a lot of features in common, so it can be helpful to concentrate on those features that differentiate one tool from another. Auto-discovery is one such feature. A monitoring tool that automatically identifies all devices on a network can reveal devices that you didn't even know existed. Auto-remediation of certain problems can save network administrators time and correct problems even when no one is staffing the administrator's console.
Simplicity is another feature to demand in network-monitoring software. The tool should display resulsts in contrasting colors and briefly describe the alert in question. The ability to click an alert for a more detailed description of the event is critical. Graphical maps of network topology help administrators get a grasp of the overall network scheme and can highlight bottlenecks or failed devices. A picture is worth a thousand words.
Monitoring tools should rely on industry standards such as SNMP (Simple Network Management Protocol) so that they are easily integrated with other software.
Monitoring tools provide two types of alerts: real-time and historical. Real-time alerts are proactive alerts regarding problems that require prompt attention, such as a server crash or a broken network link. Urgent alerts may be displayed on a centralized administrator's console; on multiple consoles; and via email, text messages or voice alerts to cell phones, PDAs and other mobile devices. Historical alerts are saved in log files, which can help chart the patterns of problems over time. For example, administrators can identify peak traffic periods during an average day. Historical alerts also provide documentation of system events, which can be used in legal proceedings against hackers.
You want a network-monitoring tool that is scalable enough to monitor all of the current and anticipated hardware on your business's network. But scalability comes at a price. Enterprise-scalable solutions tend to be very expensive, so you should avoid solutions that provide more scalability than your company will need.
Vendors license their network-monitoring software in various ways that can have significant effects on total cost of ownership. Some vendors (such as AdventNet Inc.) license software on a per-user basis, while others (such as Argent and NETIKUS.NET) require a license for each device that is monitored .If your business's network contains hundreds of devices, the latter option can become very expensive.
There are literally hundreds of network-monitoring tools available, many through the open-source community. Here is a short list of commercial vendors and their contact information.
Products: OpManager (starts at $795) and OpManager MSP (starts at $4,995)
Product: EventSentry; starts at $69 per monitored device
Network Instruments LLC
Product: Expert Observer; starts at $2,995
Numara Software Inc .
Product: Numara Network Monitor; $995 per admininstrator console
Products: Pricing for PRTG Network Monitor version 7 starts at $295 for 100 sensors
Quest Software Inc .
Product: Spotlight on Active Directory; $6.00 per user account
IT must ensure new applications are rolled out quickly, reliably, and without risk, while at the same time guaranteeing performance and availability. Read this VirtualWisdom white paper to find out how to achieve application-aligned infrastructure performance, and more. more
In an era of new technologies and cloud-based application delivery models, your business success depends on your ability to ensure optimal application performance and quality user experiences at all times. This complimentary white paper from AppNeta will enlighten you to the new frontiers in end user experience management and much more. more
Consider HP as your Microsoft Lync Solutions provider! more
Increasingly, the success of business is being tied to the network. The transformation of the network and IT can help organizations deliver and support highly available applications and services while reacting more quickly to changes in the business environment. In this complimentary white paper from IDC, learn how HP can help its customers and partners improve the overall application experience. more
This buyer's guide presents an overview of leading products on the market today and aims to improve research for companies needing to purchase or upgrade their equipment. more