Shulou(Shulou.com)06/02 Report--

This article will explain in detail what are the differences between san and nas. The editor thinks it is very practical, so I share it for you as a reference. I hope you can get something after reading this article.

Differences: 1. SAN can be regarded as a disk on the network, while NAS can be regarded as a file system on the network. 2. NAS connects the storage media in the form of file + LAN, while SAN connects the storage media in the form of block + optical fiber.

SAN: STORAGE AREA NETWORK storage area network

NAS: NETWORK ATTACHED STORAGE network attached storage

NAS is not necessarily a disk array, an ordinary host can make NAS, as long as it has its own disk and file system, and provides an interface to access its file system (such as NFS,CIFS, etc.), it is a NAS. The commonly used windows file sharing server is a NAS device that uses CIFS as the calling interface protocol. Generally speaking, NAS is actually a shared server on Ethernet using network file systems such as NFS,CIFS. As for whether there will be a file provider on the FC network, that is, NAS on the FC network, we will talk about it in the future.

Notes:

NFS (NETWORK FILE SYSTEM) is suitable for LINUX&UNIX system

CIFS (Common Internet FILE SYSTEM) is suitable for windows system

The difference between SAN\ NAS:

You can compare it this way: SAN is a disk on a network; NAS is a file system on a network. In fact, according to the definition of SAN, SAN actually refers to a network, but this network contains a variety of elements, such as host, adapter, network switch, disk array front end, disk array back end, disk and so on. For a long time, it has been customary to use SAN to refer to FC, especially the remote disk.

So, once a NAS based on FC network is designed, what should be called SAN at this time?

So, when talking about the difference between the two, we use an analogy, that is, the disk on the FC network is called SAN, and the file system on the Ethernet network is called NAS. We can understand it so simply.

NAS connects to the storage media as a file + LAN

SAN connects to the storage media in the form of block + optical fiber.

There are two physical conditions that ordinary desktops can also act as NAS,NAS:

First, no matter what method is used, NAS must have access to the questionnaire or physical disk

Second, NAS must have the ability to access Ethernet, that is, it must have an Ethernet card.

Performance comparison of SAN\ NAS:

1. SAN fast or NAS fast

First, take a look at the path map of SAN and NAS, as follows:

Obviously, the path of the NAS architecture uses Ethernet and TCP/IP protocols instead of memory when communicating between the virtual directory layer and the file system layer, which not only adds a lot of CPU instruction cycles (TCP/IP logic and Ethernet card drivers), but also uses vulgar transmission media (memory speed is much faster than Ethernet).

In SAN mode, there is one more FC access process in the path than NAS mode, but most of the logic of FC is completed by the hardware on the adapter card, which does not increase much CPU overhead, and the FC access speed is higher than Ethernet, so it is easy to conclude that if the back-end disk does not have a bottleneck, then unless NAS uses a network way faster than memory to communicate with the host, its speed will never exceed the SAN architecture.

But if the back-end disk has a bottleneck, the performance degradation caused by NAS replacing memory with a network can be ignored. For example, in an environment where a large number of mindful small pieces of Icano and the cache hit rate are extremely low, the bottleneck of the back-end disk system reaches the maximum, and the front-end Iswap O instruction will be in a waiting state, so no matter how fast the first segment of the path is, it won't help.

At this time, the NAS system is not slower than SAN, but may have higher performance than SAN because of its optimized concurrent Imax O design and file-based access rather than cluster block access.

Since NAS is generally no faster than SAN, why let NAS be born? Since NAS is not as fast as SAN, why does it still exist? The specific reasons are as follows:

The cost of NAS is much lower than SAN. The front end only uses the Ethernet interface, and the cost of the FC adapter card and switch is very high compared to the Ethernet card and switch.

NAS can resolve the CPU and memory resources on the host server. NAS is suitable for cpu-intensive application environments.

Due to the use of Ethernet, NAS is highly scalable and easy to deploy.

NAS devices generally provide multiple protocols to access data, while SAN can only access data using the SCSI protocol.

NAS can provide shared access to multiple clients on a disk array, including accessing a directory or file at the same time. In SAN mode, unless all clients have installed special cluster management software, a lun cannot be shared, and forced sharing will damage the data.

The specially optimized NAS system can process a large number of client requests concurrently, providing a more convenient access method than SAN.

Multiple hosts can connect directories on the NFS at the same time, which is equivalent to reducing the processing flow of the file system in the whole system, transforming the original multiple parallel processing into a single instance on NFS, and simplifying the system redundancy.

2. SAN or NAS is better.

Instructions on IO dense and CPU dense are as follows.

CPU intensive: the internal logic of the program is complex, and the disk access is not high.

IO-intensive: the internal logic of the program is not complex, does not cost much CPU, but can access the data on the hard disk at any time.

Both IO and CPU are dense: they are not suitable for stand-alone machines and must be clustered.

Obviously, NAS is much slower than SAN for large sequential IO-intensive environments, because the overall difference becomes apparent after a large amount of IO accumulation. However, if you want to use 10G Ethernet, you have to choose NAS, because the speed of the underlying link is, after all, the fundamental bottleneck of NAS.

In addition, NAS will exhibit strong relative performance if it is a highly concurrent random small block I / O environment or a shared access file environment. If the file system on the SAN host is fragmented, random small chunks of IO will be generated when reading and writing a file, while NAS's own file system will have a lot of optimizations with relatively few fragments. CPU-intensive ones should consider using NAS.

What are the differences between san and nas to share here, I hope that the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

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.