Data Center Designing for Cloud Computing: An Overview

Cloud computing has transformed the way businesses operate by providing on-demand access to computing resources and enabling scalable, flexible, and cost-effective IT solutions. To take advantage of the benefits of cloud computing, businesses need a data center that is designed specifically for cloud computing.

Designing a data center for cloud computing requires a flexible, redundant, secure, energy-efficient, and highly automated infrastructure that can scale up and down quickly and efficiently. By following these best practices, businesses can take full advantage of the benefits of cloud computing and achieve their IT goals.

Scalability: Cloud computing is all about scalability. A data center designed for cloud computing should be able to scale resources up and down quickly and efficiently to meet changing business needs. This requires a flexible infrastructure that can accommodate a wide range of workloads and applications.

Redundancy: Cloud computing relies on high availability and fault tolerance. A data center designed for cloud computing should have redundant systems in place to ensure that there is no single point of failure. This includes redundant power and cooling systems, network infrastructure, and storage.

Security: Security is a top priority for cloud computing. A data center designed for cloud computing should have robust security measures in place to protect against cyber threats and ensure data privacy. This includes physical security measures such as access controls and surveillance cameras, as well as network security measures such as firewalls and intrusion detection systems.

Energy Efficiency: Cloud computing can be resource-intensive, so energy efficiency is essential to keep costs down and minimize environmental impact. A data center designed for cloud computing should be optimized for energy efficiency by using efficient cooling systems, energy-efficient servers, and virtualization technologies.

Automation: Cloud computing requires a high degree of automation to manage resources effectively. A data center designed for cloud computing should have automation tools in place to streamline the provisioning, deployment, and management of resources.

Cloud computing and internet exchanges - a good mix

Cloud computing and Internet exchanges have a strong relationship as both play critical roles in delivering reliable and efficient Internet services.

Internet exchanges (IXs) are physical locations where multiple Internet Service Providers (ISPs), content providers, and other network operators connect and exchange traffic. IXs allow these networks to connect directly to one another, which improves network performance, reduces latency, and lowers costs.

Cloud computing providers, on the other hand, deliver computing services over the Internet. These services include infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS). Cloud computing providers operate large data centers that house thousands of servers, storage devices, and networking equipment to deliver these services.

Cloud computing providers benefit from connecting to Internet exchanges in several ways:

Improved Network Performance: Connecting to an Internet exchange allows cloud computing providers to exchange traffic directly with other networks, reducing latency and improving network performance.

Cost Reduction: By connecting to an Internet exchange, cloud computing providers can reduce the amount of Internet transit they need to purchase from ISPs, which can significantly reduce their costs.

Redundancy: Connecting to multiple Internet exchanges can provide cloud computing providers with redundant paths to their customers, ensuring high availability and reliability.

Access to New Customers: By connecting to an Internet exchange, cloud computing providers can gain access to new customers and markets, which can help them to grow their business.

Cloud helps low carbon emission

Cloud computing power saving strategies are implemented to reduce the amount of energy consumed by cloud computing data centers. These strategies help to minimize carbon emissions, reduce energy costs, and improve the sustainability of cloud computing.

Cloud computing power saving strategies include virtualization, dynamic resource allocation, energy-efficient hardware, data center optimization, and renewable energy. These strategies can help to reduce energy consumption, improve sustainability, and reduce costs for cloud computing providers and their customers.

Virtualization: Virtualization is a key technology used in cloud computing that enables multiple virtual machines to run on a single physical server. By consolidating multiple workloads onto a single server, virtualization can significantly reduce the number of physical servers needed, which in turn reduces energy consumption.

Dynamic Resource Allocation: Dynamic resource allocation is a technique that enables cloud computing providers to allocate resources based on the current demand. This ensures that only the required amount of resources are used, reducing energy consumption and costs.

Energy-Efficient Hardware: Cloud computing providers can choose to use energy-efficient hardware such as low-power processors, solid-state drives, and energy-efficient cooling systems. This can help to reduce energy consumption and carbon emissions.

Data Center Optimization: Data center optimization strategies such as proper cooling, airflow management, and efficient power distribution can help to reduce energy consumption and improve the sustainability of cloud computing.

Renewable Energy: Cloud computing providers can choose to use renewable energy sources such as solar, wind, or hydro power to power their data centers. This can significantly reduce carbon emissions and improve the sustainability of cloud computing.

Virtualization Models

Virtualization technologies can be hold under three types:

Type 1 Virtualization: This type of Virtualization is latest technology at Virtualization and at this model all systems are managed at a virtulization stack. Hypervisor systems are also part of this model. Mainly this model, works as a layer between hardware and guest os.

Type 2 Virtualization: This model needs a operating system installation and a working vmm over this operating system. One common problem is when there is a maintenance issue happens on the os, every guest should be restarted. Typical examples for this virtualization are Java Virtual Machine and .Net Common Language Runtime.

Hybrid Virtualization: This model works as all guest systems talk with the vmm and host os. This is also known as traditional virtualization. Some samples are Virtual PC and Virtual Server.

Virtualization for Green IT

One of the most dangerous affect of a datacenter is energy usage and unfortunately exhaust of hot air. When design our systems we do not even think how to save energy. We are just focusing to finish project. Nowadays, we are discussing energy savings for two reasons: Global economic crisis and saving the world.
As we just add a server for any application we are just talking about adding maybe hundreds of servers. When you think of the functioning times of these servers you will realize that some of your servers are working quite hard as some of them are resting. What we really get is, we use our cpus less than %20 and this causes purchasing lots of full computers. Instead we can use one hardware and share this hardware among the others. At this point, we have the chance of working on one hardware multiple servers and use same hardware utilized 24 hours.
Virtualization takes active role when we try use same hardware for multiple servers. Main function of virtualization is downsizing the energy costs, downtimes and time to deploy new servers. We will discuss virtualization technologies in detail.

Common and inter-system ground noises

In this entry we will talk about ground noises. First of all, what is ground? As a simple way to tell it is the 3rd line when you have a single-phase wiring. 2 lines are connected to your electricity system(ups, generator or city lines) Ok, then why are we connecting a 3rd line? As an unwanted situation, there is a disturbance between the power neutral(generally, blue or white wire) and the gorunding conductor(should be green). These disturbances exist as a result of noise injection into the neutral or grounding wires, wiring faults or overloaded power circuits. One common mistake is while your electrician cabling your system, he can use inappropriate neutral cable size. This also leads to electricity flow to return from ground.
However, Inter-system ground noise is different than common ground noise. Assume that you have two ups systems. First group is grounded by your IT team and the second system is grounded by your building system. When there is a difference in the ground noise between these systems, computers are impacted when they are connected to these different systems. It is also possible to pass noise from high ground noise to lower one with data cables. This situation is known as inter-connect ground noises.
The best way to stop these, all systems should be connected one ground system and connected systems will also have no problem within one ground loop. Additional systems grounding should be added as a tree structure.

Cooling Basics

As a datacenter or computer room owner you should check your expenses about your cooling requirements. On your computers whitepapers, every system has how much electricity uses. So be sure to check your total electricity expenses.
Once you find your total expense in kWatts, you should need to add computer rooms cooling requirements. As a best practice, you can add 400 btu per metresquare area in a average of 3,5 m. height. Ok?
Let's find your total requirement. Multiply your electricity requirement in kWatts for your servers by 3414 and add area multiplied by 400. This number will give you the exact BTU requirement.
As a best practice 59 kWatts cooling capacity system has an expense of 17 kWatts cooling and 5 kWatts for humidity organization. So for finding total electricity expense with cooling, you can assume that for each 59 kWatts server electricity expense you should add 22 kWatts cooling expense if you use CRAC.
We will talk about alternative ways to decrease cooling bills. Headlines will be:
- Hot air/Cold air aisle
- Usage of free cooling systems
- Exhausting hot air as it gets out of the systems
- Usage fresh air from outside if it is tempreature is less than inside
- Having underfloor airflow barriers
Coming soon...