The increasing use of platforms such as cloud, virtualization, big data or IT consumerization is driving data center demand. However, data center upgrade is a very complex, expensive and time consuming project. This is why many small data center operators do not want to have an upgrade. In fact, the efficient running of the data center and economic benefits of business are closely related. As the saying goes, the survival of the fittest, you should consider data center renovation proactively instead of passiveness. This paper will discuss some of the warning signs which may occur in your data center, and what you can do to fix each issue.
One of the significant feature of data center is the large amount of equipments and devices. And these devices and equipments generate heat when they work for a long time, just like a a large steamer, particularly in summer. Data center operators turn to the traditional fight of maintaining consistent cooling for theri IT infrastructure.
A recognized best approach to data center cooling is to isolate cold aisles from hot aisles. When containment is not working properly or new, high powered IT infrastructure is installed in an existing facility (e.g. blade server infrasturcture), the peak hot temperature can run beyond IT equipment manufacturer specifications.
While hot aisles should be hot, cold aisles should not be cold (rather, the delta or difference between hot and cold is more critical). Above all, cold aisles should not experience excessive input temperatures (77°F, namely 25°C) across the entire face of the servers and cabinets under all operating conditions. If you are living with this problem, it may be time to look at computational fluid dynamics (CFD) analysis to gauge how best to arrange cooling system and, where necessary, deploy hot and cold aisle containment. CFD analysis allows data center designers to pinpoint hot spots within a facility and concentrate on eradicating them during peak summer months. This eliminates the need for temporary cooling solutions such as standing floor fans.
When you begin to feel too busy to have a vacation, this do not simply mean that you are important for the work. Here, to your data center, it means your IT and facilities staff are spending countless hours on mundane tasks to keep computing facility infrastructure up and running, rather than focusing on ways to help advance the goals of the business.
Considering outsourcing some of these functions to a managed services provider. Remote monitoring and management have become accepted practices in most enterprises as a way to offload desktop support and allow internal IT teams the ability to focus on mission critical core infrastructure which, if down, impacts far larger groups of employees and customers. The same holds true for these components of the enterprise techinology infrastructure can also help to plan and design upgrades that, when implemented, can reduce operating costs and pay for themselves in months.
When you find troubleshooting cabling issues is an impossible task, it is time to consider new approaches to improving cabling design and deployment. With good cable management you can easily define your cable elements, setup the connectivity paths between them, pinpoint how faults in one cable and so on. This will save time, money, and improve overall reliability of the data center. If you don’t have any idea to achieve good cabling, to hire a professional distribution design team for your service.
When there is no space to deploy a additional gear or another watt of power in your data center, it’s time to consider a renovation. Like cleaning your basement or attic at home, a renovation will free up rack u-space. Under-utilized or obsolete equipment should be identified and discarded. Additionally, renovation can be used to increase power capacity or deploy new cooling methodologies for high draw and hot blade servers. Experienced data center designers begin with an analysis of key design criteria —basically IT’s requirements of the facility supporting infrastructure. This typically includes projections on growth and space, power density per rack, and cooling requirements over the next 3-5 years.
If you are not really sure how well your disaster recovery plan will protect your data in the event of a flooded data center, an earthquake, or any other emergency scenario, then your plan is outdated.
The process of a formal data center renovation will address redundancy in power and cooling, as well as begin the process of designing your data center to overcome equipment failure. In addition, offsite backup and disaster recovery, either in another of your facilities or at a cloud or co-location provider location, brings the ability to overcome a total outage due to power or disaster in your production facility.
Take a serious look at virtualization and data center infrastructure management (DCIM) tools. These approaches may require renovation to certain aspects of an older data center, but can be worth it if you gain visibility and centralized control within the environment. DCIM tools come in many flavors and feature sets which are scalable for a smaller data center. At a minimum, environmental monitoring, circuit monitoring, change control, and managed maintenance approaches can be implemented when your data center facility support assets are upgraded.
We are looking forward to green data center with its benefits. Considering green data center design techniques and imporved performance of facility supporting infrastructure removes inefficiencies in how much power goes to IT equipment versus non-IT equipment. The result can be reduced operating expenses direct to the bottom line.
Data center designers perform energy usage assessments by performing spot measurements and calculations of a site’s power usage effectiveness (PUE) and Data Center Infrastructure Efficiency (DCIE) ratios. These measurements allow designers to compare the data center against industry standards to estimate the energy efficiency of the facility. In fact, the assessment of these ratios can determine whether the potential exists for substantial cost savings derived from imporved power and cooling systems design. Once understood, a thorough evaluation of existing facility supporting infrastructure can deliver bottom line profit improvements.
Just like the clothes fashion, technologies also have fashion. With the increasing development of many new technologies, the demands on data center upgrade are also increasing. For instance, many data center managers are looking to upgrade their infrastructure, but worry their outdated data center cannot handle the upgraded infrastructure. As discussed, data center designs hinge upon key design criteria to support the IT load and the introduction of blade servers into an existing environment may disrupt the current load requirements.
One approach to overcome this problem is to implement a hybrid approach for facility supporting infrastructure. Rather than upgrading the entire data center to support the modified key design criteria, a data center cooling design consultant might recommend the need to make changes to the overall cooling approach. This may not mean a “rip-and replace” of all CRAC/CRAH (Computer Room Air Conditioning / Computer Room Air Handling) units, but, perhaps, a migration to a hybrid approach where traditional perimeter cooling is utilized in tandem with a higher density cooling apprach such as in-row or pumped refrigerant.
After learning this paper, have you already have a plan for your data center upgrade? Whether you have these issues or not, you may face the renovation demands of your data center if you want to have long-term development. Thus, it is important for us to learn the data center needs and the areas of improvement. In addition, in order to achieve a smooth renovation project and successful results, exceptional planning, significant prior experience, and professional team are necessary.