Now that virtualization and the cloud are placing more emphasis on enterprise networking as opposed to processing and, to a lesser extent, storage, the health and vitality of networking infrastructure has increased its stature on the IT manager's list of priorities.
Clearly, no one wants to build greater reliance on virtual solutions unless the ability to move data from one resource to another is rock solid. But unlike other management and monitoring functions, peering into network traffic flows to assess their health often has the unfortunate side effect of slowing down network performance. The more you try to see what's going on, the worse it gets.
The key, then, is to establish a monitoring regime that can keep tabs on network conditions and sound the alarm before problems arise, without drilling so deep into traffic flows that it acts like a dam across a river.
That's the goal of a new generation of network-management developers, who are employing the latest in intelligent designs to ensure visibility into the network without impeding traffic flow. A prime example is nMetrics, which has combined an intelligent data-retrieval engine with a massively scalable database capable of keeping tabs on flow, session, application and user traffic. The system can be deployed in enterprise, MSP and even telco settings, drawing data from device, access, edge and core systems. It uses a combination of SPAN, flow variable and SNMP toolkits to provide analysis on protocol, service, application layer and payload conditions.
A high degree of network visualization is becoming more crucial as network traffic itself becomes more cumbersome with the addition of rich media files and streaming video. Brocade and Blue Coat have teamed up on an appliance to address this need, combining the Brocade ServerIron ADX application delivery controller with the CacheFlow 5000 device. The unit creates a buffer around network resources that prevents spikes caused by unanticipated traffic loads. The idea is to ensure optimum network health without hampering performance or disrupting applications sharing the same resources.
Meanwhile, open source monitoring tools are making the transition from simple add-ons to integrated systems. GroundWork Open Source recently released GroundWork Monitor Enterprise 6.3, incorporating the Cacti network solution to gather data from switches, routers and load balancers. The move not only provides a more fully functional monitoring environment, it extends the system's capabilities over larger networks, pushing GWOS into the large enterprise market.
Smaller organizations tend to gravitate toward even more integrated systems, combining network monitoring and, say, asset management and systems analysis. Spiceworks' IT Desktop 5.0 offers a full network hardware configuration suite, plus an employee-centric hardware/software view and an integrated help desk. The package runs on Windows XP, Vista, Windows 7 and Windows Server 2003, 2008 and 2008 R2. Management can be extended to Mac, Linux and Unix systems, although the system cannot be hosted on these platforms.
With the network playing such a crucial role in advanced data architectures, selection and integration of a management and monitoring stack should not be taken lightly. Most leading network platforms will have integrated management ready to go, but these may or may not provide the flexibility and tool sets to satisfy your particular needs.
In all likelihood, a networking consultant or integration specialist will be necessary to ensure both proper functionality for today and the scalability to meet your needs tomorrow.