The core mission of any modern approach to data centre management should be to provide a holistic view of the data itself along with how that data is being stored, retrieved, managed, secured and so forth. Infrastructure management, for example, should have insight into server utilization within the context of cooling and power utilization and not only track servers but assess transaction count within those same views. In order to achieve these relationships, there must be a three-layered approach to managed hosting. These three layers or pillars are monitoring, analyzing and automating.
The first layer to achieving the relationships described above for any data centre network is monitoring. Monitoring is not simply watching but visualizing activities and details across all systems, locations and other factors that the data infrastructure comprises. Monitoring can be distributed and should be redundant. These are solutions intended to provide insights into physical assets, data and processes. When thinking about this layer, remember the wise words of Einstein about not being able to solve problems with the same thinking that created them. Outside-the-box approaches are favored here.
Monitoring lets us see in real-time how much space the equipment takes up, how much energy and other resources are used, how efficient cooling is, how utilized available resources are and so forth. Our second layer lets us analyze this data in real-time, which is at the same rate we’re able to monitor it. While there is certainly great value in being able to assess IT solutions in retrospect, the agility modern businesses require demands that patterns be identified as they emerge and problems be detected before they manifest. The ability not only to detect but to react is favored here.
A modern approach to data centre management requires the ability to react, but perhaps more importantly, it requires the capability to react as soon and as efficiently as possible. This is where the automation layer comes in. It not only automates these reactions but allows for synchronization across all layers and derived aspects of the data centre network. This is particularly important today because IT solutions are often distributed across multiple IT sites. This is what allows a process or a piece of equipment or an entire silo to kick in when some other aspect of infrastructure management has failed.
This managed hosting ideal requires the software to facilitate it, and that means accounting for every level of the data centre stack: network and power; facilities; IT infrastructure; application software; and application services. In addition, responses to emergencies as well as the routine movement of resources from one asset to another must be fully automated in order to be practical both in terms of time and cost. You may find it useful to visit Carbon60 Networks for more information.