Shulou(Shulou.com)06/01 Report--

In server storage, it is divided into two camps, namely: SAN storage and NAS storage.

In the era of big data, with the rapid growth of data processing demand, traditional methods such as SAN storage or NAS storage can no longer meet the needs of business development, and centralized storage is active again. So what are the disadvantages of centralized storage and distributed storage? How should we choose in the face of the two?

Today, global data storage is exploding, and corporate and Internet data are growing at a rate of 50 per cent a year. According to Gartner, by 2020, the global data volume will reach 35ZB, equivalent to 8 billion 4TB hard drives. The change of data structure brings new challenges to the storage system. Unstructured data accounts for nearly 80% of the storage system.

With the development of the Internet, the main body of data creation has gradually shifted from enterprises to individual users, while most of the data generated by individuals are unstructured data such as pictures, documents, videos and so on. Enterprise office processes are more realized through the network, forms, bills and so on have achieved unstructured digital archiving; at the same time, structured data based on database applications still occupies an important position in the enterprise, storing a large number of core information.

With the rapid increase of data services, the traditional single SAN storage or NAS storage can no longer meet the needs of business development. SAN storage: high cost, not suitable for PB-level large-scale storage systems. Data sharing is not good enough to support multi-user file sharing. NAS storage: sharing network bandwidth and poor concurrent performance. As the system expands, performance degrades further. As a result, centralized storage is active again.

So what are the disadvantages of centralized storage and distributed storage?

Seven differences between NAS and SAN

1 、 Fabric . NAS uses TCP/IP networks, the most common of which is Ethernet. Traditional SAN usually runs on high-speed fibre Channel networks, although more SAN use IP-based fiber architecture because of the cost and complexity of fibre Channel. High performance is still a requirement for SAN, and flash-based fiber protocols help narrow the gap between fibre Channel speed and IP speed.

2. Data processing. The two storage architectures handle data differently: NAS handles file-based data, while SAN handles block data. The story is not that simple: NAS can use a global namespace, and SAN can access a dedicated SAN file system. The global namespace aggregates multiple NAS file systems to present a unified view. The SAN file system enables the server to share files. In the SAN architecture, each server maintains a dedicated non-shared LAN. The SAN file system allows servers to share data securely by providing file-level access to servers on the same LAN.

3. Agreement. NAS connects Ethernet directly to an Ethernet switch via a cable. NAS can connect to the server using a variety of protocols, including NFS, SMB/CIFS, and HTTP. On the SAN side, the server uses the SCSI protocol to communicate with SAN disk drive devices. The network is formed by using the SAS/SATA structure or mapping the mapping layer to other protocols, such as fibre Channel Protocol (FCP), mapping SCSI through fibre Channel, or mapping SCSI through TCP/IP.

4. Performance. For environments that require high-speed traffic, such as high-transaction databases and e-commerce sites, SAN has higher performance. Due to the slower file system layer of NAS, NAS usually has lower throughput and higher latency, but high-speed networks can make up for the performance loss within NAS.

5. Scalability. Entry-level and NAS devices are not very scalable, but high-end NAS systems scale to the PB level using clusters or scale-out nodes. On the contrary, scalability is the main driver of buying SAN. Its network architecture enables administrators to scale performance and capacity in extended or extended configurations.

6. Price. Although the price of the high-end NAS is higher than that of the entry-level SAN, the purchase and maintenance cost of the NAS is usually lower. NAS devices have fewer hardware and software management components than storage area networks. Administrative costs are also included in the comparative factor. On the basis of a complex stack, using FC SAN to manage SAN is more complex. The rule of thumb is to calculate annual maintenance at 10 to 20 times the purchase cost.

7. Easy to manage. In an one-on-one comparison, NAS wins the convenience of managing competitions. The device can be easily plugged into the local area network and provides a simplified management interface. SAN requires more administration time than NAS devices. Deployment usually requires physical changes to the data center, while continuous management usually requires dedicated managers. The exception for SAN is that more NAS devices do not share the common administration console.

For medium-sized companies that need SAN and NAS, they are very popular, but lack data center space and dedicated managers for separate systems. Converged SAN / NAS has a much smaller market share than obvious deployments, but shows steady growth.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.