Data Center Automation: Enabling the Agile Enterprise

Main Findings

• Data Center Automation (DCA) is becoming a necessity
  The increasing complexity of the data center and dependent systems means that tooling is now required to keep track of, and maintain the environment in a manner which involves as little human intervention as possible
  
• Freeing up expensive human resource is becoming all important
  With upwards of 70% of an organization‘s IT expenditure being spent on maintenance and fire fighting, tooling is required to drive down costs. The availability and cost of adequately skilled human resources is becoming prohibitive – and utilizing these skills for maintenance is not cost-effective. Freeing up these resources for value-add strategic IT implementation work is a necessity
  
• Discovery, Dependency mapping, Maintenance and Dynamic Provisioning are all key areas
  Many organizations have no real knowledge of how many servers and other hardware resources they have in their environment – never mind to what level of version and patch each layer of the stack may be at. DCA tooling can automate each of these tasks, and ensure that systems are maintained at necessary critical patch levels and optimum operating levels
  
• Utility computing will drive the need for greater flexibility in the data center
  Service Oriented Architecture (SOA), Software as a Service (SaaS) and other utility concepts such as grid computing will increase the need for a fully dynamically managed infrastructure. Manually managing and maintaining such environments should not be considered – DCA provides the best way to create a managed system.
  

Conclusion
The opposing forces of greater demands from the business for increased flexibility and speed of response compared with the drive for lower infrastructure costs while complexity increases means that IT departments have to look to solutions to replace the costly and generally inefficient human aspect of managing the data centre and its dependent systems. DCA provides the way to create a highly dynamic and responsive capability that provides increased support for the business.

Background
Organizational computing and data needs have moved in a cyclical basis – in the age of the mainframe, everything to do with the computing environment was held within one environment, the data centre. Then there was the move to distributed computing, with departmental computers and data being held more locally to the users. However, as data volumes grew and the organisation became more dependent on the IT infrastructure, issues caused by lack of easy access to disparate data sources meant that the capability to deal with data or infrastructure loss was impaired. This led to a move back to monolithic data centres, with massive data warehouses holding aggregated data alongside racks of servers dedicated to specific tasks. However, different enterprise applications required different data stores, and silos of data began to appear within these data centres, and the speed of response when interrogating and reporting against these data stores became such an issue that departments started to again look to localised computing as a means of supporting their own needs. Now, the move is towards a service oriented architecture (SOA), complete with data federation (the capacity to bring multiple different data sources together for reporting and business intelligence purposes) and master data models (the capacity to have a single view of the truth for common data through a single database for information such as customer records). The increase in interoperability due to greater standards adoption within the vendor community, combined with the growth in understanding and adoption of technologies such as server and storage virtualisation is making such moves far more cost effective and attractive than has previously been the case. It is unlikely that we will see the data centre of old reappear – although many organisations will have a large centralised environment where a greater part of their computing environment will reside, there will always be data and servers that are held external to this environment – and yet this data still has to be included in any data management and reporting approach. To this end, organisations will need to invest in the tools required to ensure that a highly flexible data and infrastructural management environment is created for the future. This document looks at the needs of the organisation around existing and future computing and data needs, and at possible approaches to ensure that these needs are met in a flexible and effective manner.