Five-step planning for storage virtualization

In recent years, the explosive growth of Business data volumes (mail data, applications, business data, special confidential data, and so on) has been accompanied by an increase in storage capacity and data management difficulties, coupled with the growing popularity of virtualization, which will face an unprecedented change in the enterprise storage environment The importance of storage virtualization is becoming increasingly important at this time; the most immediate benefit of storage virtualization is to make shared storage more convenient, significantly less storage capacity, and more efficient storage management. While the heat of storage virtualization may be too exaggerated by some vendors to achieve publicity, but storage virtualization does have some real benefits for users, and here's a look at some of the basics of storage virtualization, hoping to provide some useful help to IT expert web users when they understand storage virtualization.

What is storage virtualization

The most popular understanding of storage virtualization (Storagevirtualization) is the abstraction of storage hardware resources. This virtualization allows users to isolate a large amount of physical features from storage resources, as if we were to store or extract items in a warehouse, as long as we were dealing with the warehouse keeper without having to worry about where our items were stored in the warehouse. For users, virtualized storage resources are like a huge "storage pool" where users do not see specific disks, tapes, or what path their data passes through to which particular storage device.

Storage virtualization reduces the configuration and management tasks of physical storage devices, while also leveraging existing storage resources. Without storage virtualization, it is only possible to have a single management of the physical storage device, which is certainly difficult to manage and can easily waste storage resources.

In-band and Out-of-band

The Out-of-band (Out-of-band) virtual engine is a virtual function implemented on a server outside the data path, that is, the control data and storage data are arranged on different data paths. Out-of-band can avoid some problems with In-band, but each server must install virtualized client software, which separates data paths from control paths, ensures that virtualized devices are not a bottleneck in data transmission, reduces traffic in the storage Data network, and helps improve system performance, But because of the general need to install specialized software, also vulnerable to attack.

In-band (In-band) virtual engine, which implements virtual storage within the application server and stored data paths, controls the data and the actual data needed to be stored in the same data path. In-band Virtual storage has a strong ability to work together, and it is easy to control through centralized management interface. However, In-band virtualization is fragile, either on a device or on a switch, because the In-band device is now a gateway that must pass between the server and storage resources, and the failure of the device can cause problems with the entire SAN data access. At the same time, in-band storage consumes more data network bandwidth to transmit control data, so it is easy to generate performance bottlenecks between servers and storage devices.

How users should face storage virtualization

First, plan the enterprise's storage virtualization strategy

A good storage virtualization strategy will result in a multiplier effect, so as a storage manager, you must first develop a good storage virtualization strategy, such as taking into account the need for virtualization policies? Before the implementation of virtualization, what storage resources does the enterprise have, and what resources does the enterprise need after implementing virtualization? After implementing virtualization, how does the enterprise manage daily backup, data protection, and archiving of disaster recovery tasks, and what is the difference between this and previous management? After implementing virtualization, how many data centers does the enterprise intend to have, and what is the function of each data?

There is no doubt that storage virtualization can lead to lower storage management complexity and lower storage costs, but deciding when to implement storage virtualization is a key factor, and also very difficult. Because it is important to evaluate the business that is related to the current and storage of the enterprise. Before virtualization, storage systems may be in a decentralized, unmanageable state, and after virtualization, storage managers can consolidate multiple storage systems into one network environment and manage the network environment in a unified manner. How key businesses in many businesses face the new storage environment requires storage managers to think hard.

I have communicated with some of the company's information centers and found that the domestic enterprises in the planning of storage virtualization strategy is indeed inadequate. In the growing popularity of virtualization today, or hope that we can do from the first step, a good start equals half of success, planning the enterprise's storage virtualization strategy is indeed able to bring many benefits to the enterprise IT architecture transformation.

Second, the enterprise business data for rational classification planning

You might think it would be funny to see this, and you would think that the data center of the enterprise data is of course very important and critical, according to their key Chengdu to classify data is difficult and unnecessary. The idea of such a knife is no doubt naïve, practice shows that many companies in this link error. Companies need to consider what data and how much data you can afford to lose when implementing virtualization. For example, the current storage virtualization technology is not enough to support the database, how you do not pay attention to these structured data types, the consequences can be disastrous. Storage virtualization allows you to store critical data by using those low performance storage technologies.

Therefore, it is very important and necessary to classify and plan the business data reasonably. Reasonable classification can make your storage virtualization more efficient, management is also more convenient and fast.

Third, understand the data lifecycle

Another important thing for a storage manager is the need to be familiar with the day-to-day business of the enterprise, particularly the need for a more in-depth understanding of the lifecycle of the company's data. If you don't have a concept as a storage Manager, you need to think about how much you know about your business's day-to-day business and enterprise data lifecycle.

How do you know what time to file and when to back up when you don't know the company's business data?

Assuming that you understand the business data, you know when to put the data on the right storage type and to plan for tiered storage of the data.

As a storage manager, the storage optimization strategy for the data cycle is required. There is no doubt that storage virtualization has a great advantage in this area.

With storage virtualization, storage managers can talk about all the different storage systems and their associated volumes for unified management through logical volumes. This makes it easy for users to migrate data from one storage layer to another.