Network monitoring getting more intelligent

Network monitoring systems, designed to keep the servers, storage, applications, databases, links, desktops, devices, hubs, switches, routers – the whole IT infrastructure bit – up and running with maximum availability should be at the heart of any company’s IT, SMEs included. For too many it’s not, and the cost is in IT ‘fire fighting’ and wasted resources – of IT and of users. There are three issues. First, for many, network monitoring tends to be in the heads of the people who look after the network. Second, for most others it’s in systems that speak the language of IT, not that of the user. Think of the benchmark in this area – systems like HP Open View, Computer Associates and now those from Tivoli, Micromuse and so on. Third, where network management systems are deployed, many only cover resources down to a certain level. That problem has been overcome famously by Computer Associates with its ‘intelligent agent’ technology in which smart monitoring software is embedded in every device and system to do the alerting automatically and universally. CA’s charismatic managing director Tony Martin puts the company ethos thus: “From our point of view [the infrastructure is] all points of failure.” And hence the breadth of its monitoring and automation systems which go out to full enterprise class IT software and hardware asset management, including automated software delivery, upgrades, refreshes – the lot. But dealing with the language issue, there’s now an alternative system from Mutiny, a joint venture between Manchester University and a web development company, which developed software behind the Toshiba plug-in network monitor appliance. Andy Murray, Mutiny’s technical director, says his system looks at the network from the users’ perspective. “It’s not about ‘is the router working?’. Users just say ‘I can’t print’, or ‘I can’t access the network’, which may involve several IT resources – servers, disk, CPU, memory and so on.” Mutiny is set up to monitor infrastructure operation against defined service levels, providing the network management team with a web-based high level traffic light view and the ability to drill down, while also providing alerts with email and SMS workflow and so on. To that extent, like all the others it provides people and systems with early indication of system degradation (disks getting full, the network becoming overloaded) without information overload and before it escalates to real problems. But being service level-orientated, it also maps ‘user speak’ to network resources. The result, faster problem resolution, according to Murray.