Debunk 9 Big lies about cloud computing

The hype of the manufacturers, and the self-deception of the IT people, will soon make people feel disappointed with some technology. If you are considering cloud computing strategy, please do not be fooled by these false promises.

Now, wherever you go, someone will tell you or sell you something about cloud computing. You can only understand whether cloud computing is really right for your business, and you only have to peel off the cloak of its lies. Here's what we're going to get out of the 9 big stories about cloud computing.

Lie 1: There is only one kind of "cloud". There are at least three forms of "cloud", each with different interests and risks. They are 1 "infrastructure as a service" (IaaS, providing on-demand virtual servers, such as Amazon's EC2, 2) Web services or "Platform as Services" (PaaS), providing APIs or development platforms for customers to create their own applications in the cloud, 3 software as a service (SaaS) , such as Salesforce.com CRM software.

What kind of application you're running and what data you generate will have a completely different outcome on whether or not you're using cloud computing.

Lie 2: All you need is a credit card. If you're a lone developer and have plenty of time, it's OK to configure the virtual server from the command line, but if you're in charge of a business, how do you install the operating system, multiple applications, and database connections to generate returns? is a problem you have to consider. If your business is large enough, there are many criteria for security, data format, or data quality.

Do not believe that some manufacturers say, "As long as the vendors to buy a development server, the past IT departments need 3, 4 days to configure the job now 15 minutes can be done." Michael Kollar, chief architect of Siemens IT Solutions and services North America, says his company has about 2500 servers that need virtualization to provide cloud services to users inside and outside the enterprise, but cloud services are neither safe nor suitable for company standards. cannot be integrated into a broader IT environment.

Many IaaS vendors can not even adapt to the needs of the enterprise application. Sitemasher's founder and CTO Phil Calvin tried to find a cloud service provider to manage the servers he now manages. And the result? "We can't find a vendor that can scale our standard servers on demand," he said. Neither vendor can provide the low latency performance he needs, nor does the vendor provide global load balancing across the data center.

Lie 3: Cloud computing can reduce workloads. This may be the case in the long run, but at the outset you must first consider which model of cloud computing is right for you, what applications or services are appropriate for the cloud, and how to secure appropriate security, compliance, and uptime. Also remember that it takes extra time to monitor the service ability of any one vendor.

"When you run productive applications, there are a lot of questions to consider, such as redundancy, reliability, performance, and latency," said Rightscale's CTO and founder Thorsten von Eicken. Before migrating applications to the cloud, customers must ensure that these requirements are met.

In addition, not all applications are suitable for cloud computing. Depending on the application of the server cluster, for example, it is not appropriate to share a cloud of computing resources with others, as they require that each server's configuration is not arbitrarily changed, and that there is a need for dedicated bandwidth and so on.

Lie 4: You can seamlessly integrate private and public clouds. Some cloud service preachers have presented us with two wonderful worlds: You can have the control provided by the internal data center and enjoy the low cost and flexibility provided by the external cloud, and all kinds of applications, storage and servers can be dragged and dropped as you want.

However, it is quite difficult to achieve this. For now, it takes a lot of legwork, a lot of manual reconfiguration, and a lot of engineering effort to migrate applications to both public and private clouds. Seamless integration is easier to implement only when customers run the same platform in both public and private clouds.

The key, says Siemens ' Kollar, is the need for a security infrastructure that spans two environments; a safe and economical way to replicate data or access data across two clouds.

Lie 5: There is no way to seamlessly integrate public and private clouds. This is also wrong. In fact, some vendors are providing seamless integration. For example, Kollar is expected to provide seamless integration of two kinds of cloud to its customers within 12-18 months. Rightscale von Eiken recommends some standardized configurations, data models, and automated deployment strategies that apply to both public and private clouds. When technology, standards, and processing processes are perfect, it should be possible to build a common and private resource that can be shared.

Lie 6: How can cloud computing save costs. And a controversial white paper recently published by  claims that it is only possible to save costs when running certain special platforms in the cloud, such as Linux.

Google has objected. Rajen Sheth, senior product manager at Google Apps, argues that report only considers the cost savings of low-cost servers in highly redundant architectures, which are not comprehensive. He believes that customers can save costs by leveraging "the same scalable application servers and databases" without having to buy, install, maintain, and extend their own databases and application servers. has not responded.

There are also users who point out that in the existing licensing and technical support models, customers may have to deploy software in the cloud more than they would in the enterprise, and need to pay higher fees to the software vendors.

Lie 7: Cloud vendors can ensure cloud security. Even if the cloud vendor has a number of security certifications, there is no guarantee that your specific servers, applications, and networks are secure.

Taking Siemens as an example, to ensure that the cloud environment of Siemens is safe, it needs to be "from the outside" to carefully examine IT systems to ensure that every conceivable user has access to critical information that is safe, Kollar said. Ensuring that every platform is secure is not a big challenge, but it is a huge challenge to ensure that all technologies that require security work together.

Lie 8: Just running a virtual machine is doing cloud computing. Virtualization is the creation of logical servers or logical stores between multiple physical devices, and it is just a subset of the elements needed for cloud computing. But even having a virtual machine doesn't mean it has a cloud. To fully gain the benefits of virtualization, IT departments or cloud vendors must also provide the ability to scale on demand, pay-for pricing mechanisms, and make it easy for users to configure new servers and storage devices to suit their needs.

Allowing users to order virtual servers, especially those that are reconfigured for certain special tasks, is a key savings goal for cloud customers. But just because you run the infrastructure 3 for example VMware, self service automatically happens and that's not necessarily true. Kollar said Siemens has invested "a large sum of money" in the development of a standard service menu for virtual servers and related services, and users can order these services from their private cloud as needed.

Lie 9: Cloud computing is only about technology. True, technology makes cloud computing possible, but achieving cost savings and flexibility requires the right process. Virtualization is the foundation of cloud computing, and when customers migrate data and applications between various physical devices, the former is very dynamic and highly variable.

Standardized processes can be used in the cloud to improve efficiency. Kollar said that by combining the ITIL management framework with virtualization technology, Siemens has reduced IT management workloads by 25% to 35%.

The truth about the cloud: The cloud is not takeout, nor is it a magical haven for carefree computing, but a complex resource that requires deep understanding and great effort to manage properly. There is absolutely no myth here.