Cloud computing capacity planning to take shape

Capacity planning has long been a word of mouth in the IT department, but it's just that. In the era of cloud computing, this situation has not changed much. While IT pros may quickly turn their attention to cloud computing capacity planning, they will at least once realize that they are overspending on cloud computing services. Capacity planning has emerged decades ago. At that time, IT administrators supervised large, centralized systems, and these large centralized systems cost millions of dollars. Fine-tuning systems and pre-upgrade planning are technologies that IT departments can use to minimize their spending.

But even so, the use of these tools is aimless.

"Many companies have little or no capacity planning at all," says Jean-pierre Garbani, vice president of Forrester Research.

In some cases, these skills are not within the enterprise. As technology continues to evolve, capacity planning is becoming more complex. For example, the advent of virtualization technology means that IT staff no longer need to study a large system; instead, they need to track the working status of virtualized servers running dozens of or even hundreds of of applications.

In addition, some enterprises also believe that it is not worthwhile to invest in capacity planning, because its cost input can be very large. The enterprise must procure the planning tools with multiple components and put human resources into the specific environments of the enterprise, which requires human expenditure. This process, which depends on the complexity of the enterprise IT environment, will be time-consuming and cumbersome.

As costs and complexity increase, the enterprise may simply need more cloud processing resources when necessary. In many cases, cloud computing providers position their services as a universal key to addressing traditional challenges, the challenges that it needs to face when it tries to provide adequate system resources. Vendors continue to focus on the seemingly never-ending elasticity that makes resource allocations as simple as calling or clicking a button.

While allocating resources in today's virtualized data centers is a much easier task, companies will find it a challenge to determine the modest scale of cloud computing services.

Manufacturers do make it easier for IT staff to quickly increase cloud computing resources, but there is no free lunch and convenience always comes at a price. In order to quickly provide computing resources, IT departments and cloud computing providers must be able to manage and deliver resources in a redundant manner. Of course, there must be a cost to pay for this redundant infrastructure.

In a private cloud computing environment, the enterprise must also build its IT infrastructure redundantly. At the same time, in the pricing model of the public cloud computing provider, the supplier should offer some kind of discount to the fixed users of the service. When a large number of resources are required to be used, the customer is required to pay an additional fee.

Although the resources they provide are unlimited, the limitations faced by vendors depend on the server processing resources, storage device resources, or network bandwidth resources they provide for their customers. Because the supplier uses the route model and the oversubscribed, it is necessary to include a risk factor in the planning process. For example, if half of the customers in black Friday call for more storage resources in Thursday, then cloud computing providers may not be able to meet all their requirements.

In addition, customers need to establish sufficient network connectivity between their site and the vendor's data center. This would be pointless if the vendor allocates additional capacity and the target user does not have enough bandwidth to take advantage of the increased resources. So, at the very least, businesses need to continue to plan for network capacity.

Cloud computing capacity Planning tools market

Once a business decides that it needs to track its cloud-computing resources more closely, there are other difficulties. Using tools to help determine the size of capacity in a given situation can create the problem that there are very few tools available.

"The companies that build systems and network management don't see capacity planning as a cash cow," Garbani of Forrester said.

One reason is that cloud computing capacity planning has become very complex. In addition to the traditional capacity bottlenecks of servers, storage devices, and network bandwidth, there are new issues-such as different cloud computing vendors and network connections between customer sites and cloud computing providers. Unsurprisingly, when customers are less interested in new products, the incentive for potential vendors to invest time and money in new product development is not high.

This does not mean that there is a lack of relevant planning tools in the market. The BMC software company's proactivenet configured Management suite enables analysis, forecasting, and optimization of cloud computing it resource capacity. If there is a capacity problem in the future, the system can automatically generate alerts, so that the enterprise may reduce the potential harm. The CA Company's capacity management Suite includes best-choice scenario measures so that organizations can analyze the impact of adding different resources to their cloud computing applications.

Some of the less well-known vendors also provide similar planning tools. Cirba claims that its customers have achieved 40% to 70% improvement in their cloud computing infrastructure. TeamQuest Company's products can produce customized service reporting. The Uptime software company's capacity tools enable customers to view the resource capacity they currently use in the physical and cloud computing environments.

In the coming years, market interest in such tools will gradually rise. As companies expand their business in cloud computing, their demand for efficiency monitoring will increase. As their spending on cloud computing services grows, they will want to keep a closer eye on the effectiveness of these inputs.