DCN February 2017 - Page 36

infrastrucure management

The significant growth in applications has created some massive challenges for infrastructure management , further adding to the management headache .
The cornerstone of a policy approach Policy management changes how a business defines itself . For example , a service provider may have a tiered offering and these different valueadded services may be linked to certain application requirements . The top tier ‘ premium ’ service may offer clients greater data protection and performance than the lower tier ‘ value ’ option . Note : the terminology here is not a minor point , if one were to choose between Platinum , Gold , Silver and Bronze ( the terms frequently used to describe service offerings ), one might feel somehow deficient or suboptimal . Terms such as those used in retail are much more effective ( Finest , Standard , Value ), these associate higher quality and therefore cost with the middle option being the de-facto standard that most of us end up selecting ( and not always the premium product ). The point is it forces a consumer to think about the cost and associated value and match the two .
Automation and integration play an important role in policy management , as the application owner is rarely the storage or network administrator and may not know how to configure a device for its end goal . By providing a tick box service , the users can easily choose how they want their data to be managed . By selecting pre-defined policies , they can make complex yet appropriate selections at the click of a mouse . They can choose how their data should be protected and replicated , they can decide whether it should be encrypted or deduplicated , appropriate retention limits can be set ( perhaps to comply with a local regulation ), and they can pick a performance policy to meet service level requirements specific to their application .
Abstracting the complexities of the infrastructure from the management framework enables the user to provision the application according to its requirements , with the automated programme managing the process . Picking a solution which integrates with OpenStack or VMWare is also particularly useful , as it provides applications vendors a framework from which users can choose those elements .
Predictive analytics also have an important role to play in effective policy management , by providing the insight from which users can make the policy decisions . By knowing what is deployed and how it is being utilised , the application requirements can be authoritatively identified and the changes can be made in the environment based upon those data points , rather than through the user ’ s gut feeling or on the advice of an expensive consultancy .
The requirements of applications may also change at different points in its cycle . For instance , an application may consistently exceed its normal requirements at the end of every month during payroll or consider the impact of ‘ Black Friday ’ or ‘ Cyber Monday ’ to an online retailer . At high capacity times , an app-data gap may be created , which slows down the data delivery to the application and impacts its operations .
Only when understanding that the infrastructure isn ’ t meeting the increased demand can the infrastructure management team apply appropriate measures to mediate those issues , whether that ’ s adding flash or increasing memory . When using predictive analytics , organisations can make the changes that will resolve the actual problem , rather than guessing or making ‘ blind ’ upgrades where they once again end up managing to the lowest common denominator .
Future proof against the management burden While the management headache of bespoke management may be daunting for some infrastructure managers , moving away from the inefficient and expensive lowest common denominator management process has never been more important .
The number of enterprise applications will only continue to boom as more business functions are digitised . Those organisations that don ’ t begin working towards an intelligent infrastructure management framework now will find themselves crushed by the inefficient processes and with the CFO on their back .
Adopting a policy framework approach can help avoid this nightmare , by incorporating both predictive analytics and automation to increase effectiveness , while simplifying the management burden .
36