Cloud computing expression

One for the pursuit of high-quality cloud storage solutions providers and service providers, in addressing some of the corresponding cloud storage issues, inevitably have to deal with some of the tasks on cloud computing. This article attempts to interpret the fact that if an ambitious cloud storage provider wants to provide good cloud storage services, then all the cloud computing issues it needs to address are all challenging and belong to a rare category. Therefore, we can calculate a corresponding problem, whether it is "on the cloud" or "under the cloud," and it is understood that a term is called "command line." Expressed as follows:

Command input and output

The "command" means the execution code of the calculation task, the "input" means the input data to be processed by the calculation task, and the "output" means the output data result after the calculation task is completed. Give a specific command line example:

1.WordInFile OutFile; (With Office Word file InFile for document processing results will be stored as a file OutFile)

For one user, if the calculation is "on the cloud," the command line "command input and output" is executed on the data center server rented by the user and the user's input and output data are also stored in the data center. In the "cloud" scenario, users have to use their own computing resources to execute the command line and also store and manage their own data. For the purposes of this article, let's consider only the case of "on the cloud," and use the "cloud command line" to refer to a command line executed on the cloud.

So cloud command line describes a typical cloud computing tasks, such as Google Docs to provide such an online office software services. A good cloud storage solution / service provider will inevitably have to solve some cloud computing tasks. We can also use the cloud command line to abstract a cloud computing task that must be performed for such a cloud storage vendor.

2.DeDupip1: v1ip2: v2; (Depletion deduplication is performed on the logical volume v1 of the data center server with the IP address as the deduplication scheme DeDup, and the result is stored on the IP2 logical volume v1 of the data center server. This cloud command line describes a cloud computing task a premium cloud storage service provider must do to reduce service costs.)

3.LoadBalanceip-1 ip-2 ... ip-n; (Equilibrate storage in a cluster of servers with IP addresses ip-1, ip-2, ..., ip-n so that each server can balance the processing and storage of data Ability to optimize service quality.)

4.RAICVip-1ip-2 ... ip-n; (Redundancy of disk volume V to "low-cost redundant cloud storage array" with IP addresses ip-1, ip-2, ..., ip-n , Where RAIC = Redundant Array of Inexpensive Cloud storage service providers. These low-cost cloud storage providers because of low fees can only provide some low-end, such as may be dropped, cloud storage services, not the pursuit In order to improve service quality, Such as providing cloud storage services will not be dropped, the cloud computing tasks must be addressed.

Examples (2), (3), and (4) of the above cloud command line can also be provided with cloud data security protection. For example, with (2) plus data security protection, we can use the following command similar to Unix Pipe A combination of computing tasks: the first encrypted data to do decryption, then deduplication, and finally the results of the treatment to do encrypted storage. Here's an example of a cloud command line that has been combined with this pipeline:

5.Decrypt-passphraseip1: c1 | DeDup | Encrypt-passphraseip2: c2; (This combination of the cloud command line a little technical details, the general reader skip here anyway. Passphrase decryption command is Decrypt and Encrypt Encrypt parameters used , Which points to a key so Decrypt (Encrypt) program decrypts (encrypts) incoming (outgoing) data; the symbol "|" is called "pipe" to inform the operating system of the sequential operations Decrypt; DeDup; Encrypt).

We can also do the same for pipe groups (2) and (4), and distribute deduplication into low-cost redundant cloud storage vendor arrays. The result is reliable and low-cost. Plus (5), data can be securely secured against low-cost, reliable cloud storage services to prevent theft of user data by low-cost cloud storage providers. We can also consider the cloud data retrieval query task, but also based on security policy query retrieval, and so on, not in this list.

Why is it that we are more ambitious to say that a cloud computing task to be solved by a cloud storage vendor that is focused on providing good cloud storage services will be more than a general cloud computing task such as the cloud command line (1)? We may wish to make the following observations: cloud command line (1) a class of cloud computing tasks are usually used by individual users, to handle a specific task, is relatively simple. Therefore, due to service quality problems caused damage to the user will generally be relatively limited, preventive measures are relatively easy to achieve. For example, we can imagine the cloud command line (1) in the implementation of server downtime caused user data loss. A simple precaution can be to make Word on the cloud a frequent AutoSave.

A relatively large computing tasks can also use the more frequent CheckPoint to save the calculation of the intermediate state to reduce the loss due to downtime. Cloud computing command line (2) to (5) category, and a little bit of imagination on their combination of cloud computing tasks not only for individual users to provide high-end cloud storage services, more likely scenario is: this type The task of computing is what we commonly call Information Lifecycle Management (ILM), which is primarily a storage service provided by high-end cloud storage service providers that face the standard service enterprise or organization user. Such users generally need to provide ILM cloud storage services with massive data processing capabilities, and their corresponding speed, reliability, security, version management, query search quality, etc., must be professional. For such ILM tasks, it is not hard to conclude that cloud storage is more challenging than cloud computing, which is just a sub-issue of cloud storage.