By Michael Brandenburg, Technical Editor
The elusive goal of network monitoring tools is not only to alert administrators when there is trouble on the network but to develop trends in the health of the network and ease the burden of SLA reporting. While vendors promise the world in their network monitoring tools, the reality is that network monitoring reporting can be hard to decipher and networking teams must work to focus their tools for their specific needs.
Myth: Network monitoring solutions take the pain out of filtering or rolling together the flow of messages into usable information.
Reality: Vendors usually offer report templates and basic filtering, but it will still be up to the administrator to make the chosen solution useful.
What may really work: Every enterprise has its own unique processes and architecture, and any monitoring solution will have to...
Myth: Networking monitoring tools will support any device on a customer's network.
Reality: While this is technically true, since most solutions use SNMP (Simple Network Management Protocol) to pass along diagnostic and status information, a typical network device, such as a router or switch, can have hundreds if not thousands of SNMP messages that broadcast the current status of the device. They report issues the device may be experiencing, as well as performance numbers. With thousands of potential traps coming from each and every monitored network device, the noise level can quickly reach the point of uselessness, with the important messages getting lost in the stream of informational ones. It also would be nearly impossible for the vendors to identify and translate the messages coming from every make and model of network device on the market. So, in most cases, the vendors focus their attention on switches, routers and other devices most commonly found on enterprise networks. This no doubt eases the pain for companies running on market-leading gear but is little consolation for customers monitoring products outside the top-tier vendors.
What may really work: Most software vendors would be more than happy to bring in their own professional services staff to make their solution work with the customer's environment, but there are alternatives for the users who want to keep the work in-house. Many solutions, both open source and commercial, have built up strong online communities, enabling users to share ideas and, in many cases, report templates and configuration files. A strong, helpful user base can not only help new users get up and running faster but can also eliminate the need to continually reinvent the wheel for every new deployment.
Myth: Most network monitoring solutions offer both real-time event notification and historical data for all of the devices on the network.
Reality: Administrators need to be aware when evaluating monitoring solutions that packages actually process and archive the data collected from across the network. Many vendors have to balance the ability to update the status of devices in real time with collating and developing historical data for all devices on the network. In some cases, only a subset of monitored devices can be tracked in real time, forcing customers to decide which components of the network are considered mission critical and require a deeper level of monitoring. Similarly, network monitoring tools that focus on real-time updates and notification typically scale back the historical data.
What may really work: Understanding the needs of the organization and the role that network monitoring will play within IT is key. Deploying a tool for the help desk has a set of requirements different from those needed when presenting the CIO with a dashboard showing SLA goals. Like any technology project, setting expectations early will minimize any misunderstandings and ensure that the solution deployed matches what is needed.
This was first published in June 2010