Which cloud use case can "thin" costs?

With the rise of cloud computing, many organizations begin to evaluate cloud use cases in order to migrate applications to cloud . They are analyzing which use cases are most meaningful. But simply migrating the application to the cloud is not necessarily the best goal, says Cantara, vice president of research , "When migrating to the cloud, production is the target of specific output." While the cloud cuts the cost of capital, there is no empirical evidence that cloud computing will cut overall cost of ownership over the long term. The

She says better cloud use cases can accelerate the time of business output, reduce capital consumption, or reshape business processes, on the basis of increasing customer satisfaction and thus increasing profitability. However, the connection between performance and business output of a cloud service provider is indeed blurred. Although Gartner's research found that 66% of organizations say they manage cloud service providers based on business output, only 27% are associated with actual service level agreements (SLAs).

Clear Cloud implications

In your own datacenter, an internal private cloud deploys a cloud-based infrastructure. An external private cloud is a managed model in which external service providers specialize in hosting cloud services. However, for security and privacy reasons, you should isolate your environment from each other.

Community Cloud leverages restricted communities, such as partners or suppliers in a variety of enterprise business process networks. Public clouds are usually what people think of when they talk about the cloud, and they are related to the external cloud environment, such as Amazon Web Services, Rackspace, Azure, Expedia, and Salesforce for sales automation.

Customer relationship management applications are more likely to be used in public clouds. In contrast, applications such as enterprise resource planning gradually integrate more back-end data with higher security requirements and are most likely used in private clouds. Cantara says organizations in developing countries are more likely to use public clouds rather than private ones because they lack the infrastructure for private clouds.

But no cloud can exist in isolation. Very few are completely independent of the Enterprise center. Eventually, Cantara expects the hybrid cloud to become more common.

Cloud Maturity Expansion Opportunities

At first, cross-gather applications have challenged many enterprise applications. The main focus of using the cloud is developing and testing Web applications, rather than commercial, off-the-shelf applications, says Robert Green, chief cloud strategist at Enfinitum Consulting.

From a cost standpoint, it is advisable to transfer development and test systems because servers and instrumentation can be shut down when the developer stops working. "You don't need to run a server when developers are resting at home," Green said. Make sure you use the Automation Control Panel to reduce the time and effort to expand and reduce the cloud use cases.

"Now we've got a bunch of people migrating applications from a service perspective to software, such as applications for production (Office 365) and offline backup and archiving (Dropbox), and incremental development and testing platforms," Green explains. Many organizations use the cloud as a mainstream way to save internal and commercial off-the-shelf (COTS) applications. They do not have a large amount of internal system requirements and minimize requirements to connect to the data center and the main framework.

But as the underlying infrastructure continues to improve, Green sees more organizations moving three of layers of applications to the cloud. This application includes integration into databases, Web servers, and web-based clients.

Get performance on budget

Green says the biggest challenge is speed. Multi-tier applications tend to have a high level of input/output operations (IOPS) requirements per second and increase RAM directly attached to virtual machines (VMS). As a result, performance cannot remain stable when some applications are moved to the cloud.

If you have 10 servers running internally and migrating out, 10 servers may be cheaper than the cloud. But once you've added the entire cost structure to move the equivalent performance to the cloud, the cost will be less clear.

To address this problem, many cloud providers offer high-performance IOPS functionality as part of their products, such as solid-state Drive (SSD)-driven platforms. This can hide some performance constraints and make the cloud infrastructure more applicable to a wide range of applications.

Green shows that focusing is important for understanding performance metrics. This is the bottom line for your application to practice well in the interior before migrating to the cloud. These applications can be compared to the same applications in the cloud platform. So you can get the same kind of comparison.

Allocating performance, not just VM

Cloud providers are getting better at dealing with performance issues. They begin to understand performance issues and offer more choices. "The best part of the cloud is that as consumers start complaining and initially choose to change, providers offer more choices," says Green. ”

The main infrastructure, the service providers, such as Amazon, Azure and Rackspace, focus on price competition. But, says Green, "the only way to differentiate it is if one of them is cool, faster, or better, the others will roll the same." ”

He expects to see new focus, such as providing IOPS throughput through the right SSD instance. "They will continue to innovate to capture business because they cannot compete only on price," says Green.

If you invest the right size cloud configuration for your application needs, the economy will improve. For example, if your COTS or application server specifically requires 8GB RAM, but only 2GB in practice, you do not need to adjust the cloud machine to support 8GB. You may get 2.5GB. This reduces the cost of narrowing down the other aspects of the cloud that you might use.

Develop appropriate metrics

The current state of the bottom line process begins with identifying critical application metrics. "If you don't have these metrics to capture, you don't know if the cloud environment is right for you," says Green.

Once you've identified metrics, you need to focus on whether you can add value if you migrate to the cloud. You want to take advantage of synergies. If you are migrating, you may well improve. If you think about the state of the future, consider how to use automatic scaling and reasonable streamlining.

Also consider what you want from a budget and operational perspective. Once you have outlined these things according to the cloud use case, you will then have to migrate the process. You'll know from the first day that it works.

"You'll see a lot of people jump in without a deep analysis," Green said. Initially it was cost-saving, but then the application ran slowly because they ran more VMS than initially thought, or they had to run longer. In addition, these are usually implemented without governance, so the things that run may still fall back.

Moderately extended

With horizontal scaling, you can extend applications by adjusting a component in a three-tier application, such as going to a Web server without having to adjust the relevant application or database server. By scaling vertically, all components need to grow at the same time.

A typical Web application can be scaled horizontally, but depends on how the application is developed. On the other hand cots software tends to more vertical expansion characteristics. But these will change, so an accurate assessment of how the plan can move to the cloud is effective.

Once you determine how the system expands, you have the opportunity to configure automatic scaling more efficiently. This allows you to reduce the absolute minimum. The automatic extension tool can then grow or shrink, all depending on the infrastructure requirements on the preconfigured triggers.

Manage growth

The next step in determining the cloud use case is to maintain performance as requirements grow. These include measures for database queues, log file specifications, or CPU memory utilization.

At the enterprise level, most organizations focus on the automation layer on top of the IaaS layer, rather than the extension capabilities that can be built into the PAAs layer. In many cases, they take advantage of tools such as ServiceMesh, Scalextreme, or rightscale that can easily remotely control applications.

Today, a typical enterprise is more concerned with these tools than using the automatic extensions that have been built onto new platforms, such as Apache Stratos,green said. A good migration tool allows you to envision an IaaS platform migration from an internal infrastructure. Then once you get into the public IaaS you can use a similar scalextreme tool to manage the overall design.

These tools can also help simplify development requirements to manage workloads across platforms from a single monitoring application. "These have been built into cloud assets and then used a tool to work together for governance and access," says Green. ”

"Responsible Editor: Dan TEL: (010) 68476606"