How does the Cloud data store "complete" the most cost-effective?

The high cost of cloud data storage shattered the dream of corporate public cloud business case. Some companies say they can buy hard disks at the same cost as monthly cloud storage and use-case fees. At the same time, Microsoft and Google free to give its SkyDrive and Google Drive services. Cloud consumers may find it hard to figure out where the price difference lies. To fully understand the content, look at what you actually get when you buy cloud storage.

Disk capacity is relatively less expensive alone, especially in large-scale applications. You can spend more than 100 dollars on a little bit of money to buy an external hard disk, for example, you can fake a 100TB capacity of a large server, you can for 10,000 end users on this server storage 10GB, of course, we assume their I/O activity is limited.

On the other hand, a single user with 10GB storage and high transaction rate can easily consume an entire server, or upload a cloud provider's data center router and storage Area network (SAN). Cloud data storage costs are higher when data is accessed on a regular basis, especially outside the cloud.

The cost burden reflected by the data usage fee is imposed on the cloud administrator. Companies believe that accessing cloud data forces them to expand their connection speed in order to avoid crowding. This, in turn, increases the cost of the network.

Cloud consumers need to help manage data costs, and four strategies have been developed to help them address these cost dilemmas.

Discover the level of data storage and usage. When designing an application and associating it with the data, think of creating a storage hierarchy. While the memory data cache may look silly on the cloud, by keeping most of the memory elements accessible, it can reduce the cost by reducing the data access rate. Study how to combine different classes of cloud storage into one level-from real-time to archive, to influence data prices.

Consider local storage as a service to support cloud applications. Although block or file system I/O between local resources and the cloud cannot be cost-effective or smooth, it may be useful to store data locally and be presented to cloud hosting applications in the form of a Rdbs query-level Interface (DBaaS). This can reduce costs for cloud data and help cloud hosting applications access local data faster, and businesses can move more applications to the cloud. This strategy is most effective for large database query outputs with relatively small result sets.

Use the digest database for the cloud data store. Cloud analysis is particularly able to operate on summary data rather than business/transactional detailed documentation. Go to the store to find product sales, not registration details, through several orders of magnitude to reduce the number of records. This can reduce the cost of data storage.

Critical business applications are most difficult to deploy in the cloud because they often use large databases to record transactions. In retail applications, for example, these databases contain the market and change inventory. While many companies deploy the Web front-end for critical business applications in the cloud, we want cloud backup application components. In this case, the key is to control the cost of data, may create a multi-level application.

Consider Web retailing systems, such as. These systems typically generate most transactions for very little data. If the front end is set up with idle "deposit (deposit)" items, representing a quantity less than this quantity is actually available, the company is able to take orders without checking inventory safety. This will leave the inventory item out of the cloud storage. So the enterprise can afford to access the cloud storage database every time the cost, this system is to save money.

The application can then send updates to the local database at a certain frequency and receive an additional "allocation (allotment)" that they can undo when the order is reset. This allows most orders to be processed, even for a brief outage in the data center. Similar to the caching process, data can be used for other online activities and to save corporate money.

Save the backup copy. Consider saving copies of some of the data on some inexpensive backup sites, and then "recover" to a more expensive cloud storage as needed. Data is not available during recovery, but the impact of power outages is minimized. For databases with low-level activity, this strategy may be the best way to manage costs, while reaping the offline processing choices for events after data center failures.

Because cloud computing changes all aspects of it planning, so does the database software and architecture. Companies must pay attention to how these changes affect them, whether their virtual resources or private clouds also include their public cloud choices and prices. Storing data in the cloud can never be cheaper than storing data on a written hard disk, but the advantage of using cloud storage may give the purchaser the best cost and the highest input/output ratio.

(Responsible editor: The good of the Legacy)