Ghost in the data center: Tips for managing wireless sensor networks
In 1949, philosopher Gilbert Ryle coined the phrase, “the ghost in the machine” to describe the interdependence of the mind and the body. The knowledge we gain directly influences the actions we take.
For several years, government network administrators have tried to turn knowledge into action to keep their networks and data centers running optimally and efficiently. For instance, they have adopted automated network monitoring to better manage increasingly complex data centers.
Now, a new factor has entered this equation: wireless sensor networks. These networks are composed of spatially distributed, autonomous sensors that monitor physical or environmental conditions within data centers to detect conditions such as sound, temperature or humidity levels. Like ghosts in the data center, wireless sensors are bringing to light knowledge that can help IT administrators take action to correct problems in a timely manner.
However, wireless sensor networks can be extraordinarily complex. They are capable of providing a very large amount of data, and they can be just as detrimental to application and network performance as bandwidth bottlenecks and outdated equipment. This can make it difficult for managers to get an accurate read on the type of information their connected devices are capturing -- which in turn can throw into question the effectiveness of an agency’s network monitoring processes.
Fortunately, there are several steps federal IT managers can take to help ease the burden of managing, maintaining and improving the efficacy of their wireless sensor networks. By following these guidelines, administrators can take the knowledge they receive from their sensor arrays and make it work for their agencies.
Establish a baseline for more effective measurement and security
Before implementing wireless sensors, managers should first monitor their wireless networks to create a baseline of activity. Only with this data will teams be able to accurately determine whether or not their wireless sensor networks are delivering the desired results.
Establishing a baseline helps achieve two more specific goals. First, it allows managers to more easily identify any changes in network activity after their sensors are deployed, which, in turn, provides a true picture of network functionality. Second, a baseline provides a reference point for potential security issues. Any suspicious activity or alerts can be compared to the baseline, allowing security teams to more effectively pinpoint the problem and establish a response plan.
Set trackable metrics to monitor performance and deliver ROI
Following the baseline assessment, administrators should configure trackable metrics to help them get the most out of their wireless sensor networks. For example, bandwidth monitoring that lets managers track usage over time can help them more effectively and efficiently allocate network resources. Watching monthly usage trends can also help teams better plan for future deployments and adjust budgets accordingly.
Metrics (along with the initial baseline) also can help agencies achieve measurable results. The goal is to know specifically what is needed from devices so that teams can get the most out of their wireless sensors. Otherwise, how are they to know if their investment is paying off? With metrics in hand, managers can understand whether or not their deployments are delivering the best return on investment.
Apply appropriate network monitoring tools to keep watch over sensor arrays
Network monitoring principles should be applied to wireless sensor networks to help ensure they continue to operate effectively and securely. The good news is that many of the tools that teams are already using to monitor typical network functions can be applied to their wireless sensors.
For instance, network performance and bandwidth monitoring software can be effective at identifying potential network anomalies and problematic usage patterns. These and other tools can also be used to forecast device scalability and threshold alerts, allowing managers to act on the information that sensors are sending out.
These tools, along with the other strategies mentioned above, are designed to do one thing: provide knowledge that can be turned into effective action. Managers can use these practices to bridge the gap between the raw data that their sensors are providing and the steps needed to keep their networks and applications running. And there is nothing scary about that.
Joe Kim is executive vice president engineering and global CTO at SolarWinds.