Data centers are insatiable creatures. The more processing power and bandwidth you make available, the more resources they demand. Every improvement in data center performance is quickly consumed by the latest applications or a new project that requires even more throughput, faster processing speeds, and more data storage. Keeping pace is an ongoing challenge, but there are strategies that IT managers can adopt to stay ahead of demand.
According to TechTarget’s IT Priorities Survey, 42 percent of IT executives plan some kind of data center infrastructure upgrade in 2015. Of those surveyed, about 47 percent increased their hardware budget, 48 percent said they were allocating more money for software, 41 percent were adding money for cloud projects, and 32 percent were adding funds to increase staffing. The question becomes: Where are they investing in order to improve data center performance, and how are they planning to improve operations?
Here are just some of the areas that the experts have identified that value-added resellers can provide to help tune today’s data centers:
1. Performance monitoring – Before you can tune data center performance, you need a baseline for comparison. Performance monitoring tools are an excellent place to start. Higher-level performance monitoring tools have been available for some time, but as data centers become more sophisticated, you need to use more sophisticated monitoring tools. Data Center Infrastructure Management (DCIM) systems continuously collect performance data that can highlight weak points and potential bottlenecks using predictive analytics. Data center monitoring and asset tracking can provide real-time data about hardware and software, which can be useful in identifying areas that need to be tuned.
2. Capacity planning – In addition to performance monitoring, you need capacity planning. As new demands are placed on the data center, you have to balance hardware, software, and infrastructure capacity. New data delivery models and factors, such as cloud computing, make capacity planning more complex. IT managers not only have to look at data capacity but consider facilities management issues such as power and cooling. Any move to modify or improve data center performance should fall within the capacity plan.
3. Data center consolidation – The first step in tuning the data center is to streamline operations. Consolidating data centers is high on the list of priorities for 33 percent of enterprise IT managers. Managers are recognizing that they can plan consolidation into their routine hardware refresh cycles, reducing the physical infrastructure rather than simply replacing outdated hardware. Consolidation helps fine-tune the data center by reallocating workloads. For example, more virtual machines can be hosted on the same server rather than buying more servers, resulting in faster data exchanges and eliminating network traffic at the same time you are reducing hardware. With data center consolidation, you can consolidate management tools, streamline application performance monitoring, rebalance workloads, and generally clean house.
4. Virtualization – Implementing and updating server virtualization can improve performance in various ways. It improves internal communications between virtual machines, keeping data traffic local on the server. Network virtualization also allows IT to treat the Ethernet infrastructure as a server pool, allocating virtual machines, along with security and network services, as needed.
5. Software-defined networking (SDN) – SDN and even software-defined data centers are a new way to automate enterprise operations. SDN is powered by open standards, and it can be used to manage data traffic, optimize cloud services, and manage bandwidth for projects such as big data. It’s also designed to be centrally managed so IT executives get a logically centralized map of the entire infrastructure. SDN promises to provide a fast, flexible way to reconfigure the infrastructure from a single location. The result is better workload performance and less data traffic.
6. Cloud computing – Virtualization and SDN technology make an ideal foundation for cloud computing. Private clouds not only help consolidate data centers, but they also offer self-service and automation support. A private cloud also makes an ideal testing ground for new services and applications prior to moving those applications to be hosted less expensive public cloud services. Migrating routine tasks to the cloud frees both capacity and storage for mission-critical workloads.
7. Data Storage – Tuning data storage can improve data center performance. For high-demand, high-performance computing, consider using a solid-state drive (SSD) or flash technology in order to speed up data access. Hybrid data stores that incorporate both SSD and mechanical disk drives are becoming increasingly popular because they offer faster performance at a lower cost. In general, the closer you can store the data to the process, the faster the response time. Consider segmenting hot and cold data for use in the data center; storing hot data in high-performance, fast-access drives and using the cloud and slower, less expensive data repositories for cold data and backups.
These are just some of the elements that can have an effect on data center performance. Each factor will have different hardware and software considerations and will yield different results depending on the computing infrastructure. Therefore, no matter what data center technology you have on your line card, you can find multiple ways to tune data center performance for your customers.