Issue link: http://cp.revolio.com/i/207723
data and sensitive customer information regardless of where it resides. In particular, IT will need to pay attention to some areas that have been relatively unprotected in the past. For example, rootkit hypervisors are at risk, as well as memory BIOS. Platform purchase criteria should include security requirements that protect these vulnerabilities. cate resources to minimize operating costs. Be sure to introduce monitoring tools that offer real-time, fine-grained data gathering. For example, server inlet temperatures provide more accurate maps of your data center conditions compared to monitoring returned-air temperature at the cooling system. Some Foundational Approaches That Complement These Trends Optimize the existing data center. At a glance, the major data center trends for 2014 would seem to relate to platform decisions and server growth. Certainly the processor innovations will bring some exciting capabilities to the data center and IT teams. However, as usual, successfully introducing new platforms is just one dimension of a forward-looking strategy. While you evaluate the emerging solutions and define next year's data center initiatives, you should continue to watch the industry trends and, just as important, you can monitor and learn as much as possible about your existing data center. If you don't fully understand your starting point, and optimize current operations, the path forward will be significantly more risky. Here are some steps you can take today, and the benefits you may expect in terms of the major trends just discussed: Know your data center use patterns and activity drivers. Begin with a good picture of your starting point. Deciding where to add more compute power or which workloads to migrate to the cloud requires that you understand the current workloads and activity levels in your data center. Similarly, identifying the overloaded servers or racks or groups of racks can point out where your data center is most vulnerable to failures and security breaches. Besides monitoring server utilization rates, be sure to monitor and record power consumption and cooling demands within your data center. Energy costs are an increasing data center cost component, and should be taken into consideration when you are considering outsourcing to the cloud or introducing any new servers. Power and thermal maps of your data center, combined with energy management analytics, can help you identify hot spots (heavy usage), prevent conditions that lead to equipment failures, and intelligently allo- 16 | THE DATA CENTER JOURNAL As the data center boundaries extend out to the cloud, the management console becomes a critical tool. To accurately determine how to allocate resources and group servers, and to know when and where to add additional power, cooling, and compute capacity, IT will need a management solution that takes into account ongoing trend information for single servers and groups of servers. The management console should also give IT the ability to limit power to servers and racks. With power capping capabilities, rack densities can be increased without risking spikes that would otherwise lead to outages or shorten equipment life spans. IT can avoid under-populating racks and put off deployment of new racks with the confidence that fully loaded racks are protected Identify and address current security weaknesses. Multi-layered defense strategies tend to focus on blocking external threats, but IT can also take advantage of integrated isolation and tamper detection features that are available for servers. Talk to your technology partners about capabilities for preventing software-based attacks such as attempted breaches at the rootkit hypervisor level. Server capabilities can also help you reset attacks designed to compromise platform secrets in memory as well as BIOS and firmware update attacks. Existing platform features work cooperatively with current virtualization technology, and give IT the ability to bolster defenses through measurement, memory locking, and essentially isolating the launch-time environment. By taking advantage of hardware-based protections in the processor, chipset, and third-party trusted platform modules (TPTs), IT can better resist attacks and build more robust data center platforms. Besides lowering costs, IT can provide higher-value capabilities such as enhanced workload controls and reporting features that feed compliance dashboards. The use cases for these security features are compelling, and focus on verifying launch time components and enforcing known-good configurations of critical software. Trusted launches can verify platform integrity to lower risk from critical system malware, and thereby reduce support costs and lower the incidences of data breaches. For virtual and cloud environments, trusted pools help IT enforce control of workload assignments such as restricting sensitive VMs to trusted systems. For compliance, hardware-based reporting features can return platform trust information, to help IT establish and verify adherence to data protection and control standards. Conclusions It makes sense that a proactive IT stance calls for improving visibility and control in the data center. If you haven't already, you still have time before the New Year to investigate the relationships between energy, cooling, and compute resources and populate a data center map that you can use to optimally plan for the high-impact trends of 2014. The good news is that your planning efforts can also yield immediate benefits for your business. Real-time monitoring has been proven to enable energy cost reductions of 15 to 20% in the data center, and automating power and thermal alerts with the same resource management platforms can save IT time while driving up service availability. Automatic monitoring and logging builds an accurate foundation for trending studies for your data center, and put you in a credible position to defend capacity plans and proposals. Most important, the latest data center infrastructure management solutions put you in control and enable more intelligent policies for resource allocation and change management. No need to wait for the New Year. These are just a few of the reasons to get ahead of the 2014 data center trends as early as possible. n About the Author: Jeff is the general manager of Data Center Manager (DCM) Solutions at Intel Corporation, where his team is pioneering power- and thermal-management middleware that is delivered by data center infrastructure management (DCIM) software companies and OEMs. *Credit: Judi Barrett, children's author, who published the book with this title. www.datacenterjournal.com