The four major storage technology to analyze the advantages and disadvantages
The four major storage technology to analyze the advantages and disadvantages
The issue of data storage is very important, however, in the actual application is a series of wrong things. Often appear drop disk, volume lockup and many other problems, seriously affecting the normal use of the overall system, so data-specific storage has become one of the most concerned about the security products on the market.
Data traditional storage
In the current digital field, the most commonly used are the following four storage methods: hard disk, DAS, nas, san.
1. hard disk
Whether it is a dvr, dvs rear mounted hard disk or directly behind the server connected to the expansion cabinet, are using the hard disk for the storage method. It should be said that the use of hard disk storage, and can not be regarded as a strict sense of the storage system. The reasons for this are as follows:
First, it generally does not have a raid system, there is no redundancy protection for the data on the hard disk, and even if there is, it is realized through the host side of the raid card or soft raid. Serious impact on the overall performance;
Second, its expansion capability is extremely limited, when the video time is more than 60 days, often can not meet the storage needs of the video time;
Third, can not realize the data centralized storage, the later maintenance costs are high, especially in the dvs after the way to hang the hard disk, the maintenance cost is often more than the cost of the purchase within a year.
It should be said that hard disk storage is not suitable for large-scale digital video surveillance system applications. In particular, the need for a long time to record the digital video surveillance system. Generally this approach is coexisting with other storage methods in the same system, as a buffer for other storage methods or emergency replacement.
2. DAS (Direct Attached Storage)
DAS (directattachedstorage), all known as direct attached attached storage, the use of DAS can be very simple to realize the capacity expansion of the platform, at the same time, the data can provide a variety of rald level of protection.
When using the DAS approach. Deploy the relevant .hba card on the video storage unit. Used to establish a data channel with the back-end storage device. The front-end video storage unit can be a dvr or a video storage server. The channel can use fiber optic, ip network cable, sas cable or even usb, 1394 cable and so on.
The DAS approach does not simultaneously support many video storage service unit access at the same time, and its expansion capability is heavily dependent on the selected storage device’s own expansion capability. So in a large digital video surveillance system, the application of DAS storage will cause a great deal of difficulty in system maintenance.
It is because of these characteristics of DAS storage that this storage method is generally used for dvr expansion or small digital video surveillance projects.
3. NAS (Network Attached Storage)
NAS (networkattachedstorage). All known as network attached storage, is a professional network file storage and file backup device, or known as network direct storage device, network disk array. At the same time NAS can provide a variety of raid level protection for data.
NAS devices and multiple video storage service units are connected via ip networks, communicating in accordance with the tcp/ip protocol, and transferring data in a file i/o (input/output) manner. A NAS unit includes a core processor, a file service management tool, and one or more hard disk drives for data storage.
Using NAS can simultaneously support multiple hosts to read and write at the same time, with excellent sharing performance and scalability; at the same time, NAS can be used in complex network environments. The deployment is also very flexible.
But because NAS uses the cif/nfs protocol for file-level transmission of data, so the network overhead is very large, especially in the write data when the bandwidth utilization is generally only 20% -40%. So currently NAS is generally used in small network digital video surveillance systems or just for part of the data shared storage.
4. SAN (Storage Area Network)
SAN (storageareanetwork), known as the storage area network, through the switch and other connectivity devices will be connected to the disk array and the related servers connected to a high-speed dedicated subnet. At the same time, SAN can provide a variety of raid level protection for data.
A SAN provides a dedicated, highly reliable storage network. Allowing for independent increases in their storage capacity also makes management and centralized control (especially for when all storage devices are grouped together) more streamlined. It is because of these characteristics, SAN architecture is particularly suitable for large network digital video surveillance system storage applications, can cope with thousands, tens of thousands of front-end monitoring point of storage.
Currently, SAN is mainly divided into FC-SAN (Fiber Optic Storage Area Network) and ip-SAN (Ethernet Storage Area Network). The difference between them is the connection lines and the use of different data transfer protocols. Although FC-SAN due to the use of specialized protocols can ensure more stable and efficient transmission, but its deployment methods, construction costs are much higher than ip-SAN, so the current in the large network digital video surveillance system is more often used in ip-SAN architecture.
The Predictive Function of Big Data is the Core of Value-added Services
The Predictive Function of Big Data is the Core of Value-added Services
From the emerging industry of the Internet, which is at the forefront of the development of big data, to the traditional industries such as healthcare, electricity, and communications, which are closely related to human life, the wave of big data is changing people’s way of production and life all the time. The arrival of the big data era has brought a lot of momentum and great value to various industries at home and abroad.
The newly released report said that the global big data market size will usher in a whopping 26% compound annual growth rate in the next five years – from $14.87 billion this year to $46.34 billion in 2018. Major companies, enterprises and research organizations around the world have extensively explored and experimented with big data business models, and while there are still many models that remain unclear, some mature business models are emerging.
Two storage models dominate
Every web page, every picture, every email on the Internet, every SMS message, every phone call in the communications industry, every household’s electricity consumption data in the power industry, and so on, all of these footprints have been recorded in the form of “data” and are growing at a geometric rate! This is the most important thing that the era of big data has brought to us. This is the most intuitive impact of the big data era.
Because of the sheer volume of data and the unstructured nature of the data, many of the existing storage media and systems greatly limit the mining and development of big data. In order to better solve the problem of big data storage, domestic and foreign enterprises and research institutions have made many attempts and efforts, and continue to explore the prospects for its commercialization, and currently formed the following two relatively mature business model:
Scalable storage solutions. This storage solution helps governments and enterprises categorize and prioritize stored content and store it efficiently and securely on appropriate storage media. And traditional storage solutions with storage area networks (SANs), unified storage, and file consolidation/network-attached storage (NAS) are unable to provide and scale the flexibility needed to handle big data. Instead, the new generation of storage solution providers, represented by Intel, Oracle, Huawei, ZTE, and others, offer a full range of storage solutions suitable for large and small and medium-sized business levels to meet the needs of multiple applications of big data through standardized IT infrastructure, automated processes, and high scalability.
Cloud storage. Cloud storage is a cloud computing system centered on data storage and management, and its structural model is generally composed of four layers: storage layer, basic management, application interface and access layer. Through easy-to-use APIs, it is convenient for users to put all kinds of data inside the cloud storage, and then charge according to the usage like using water and electricity. Users don’t have to care about data storage media, network conditions, and security management, but only need to buy space from the provider on demand.
Source data value rises
In the red-hot era of big data, as data accumulates, the value of the data itself continues to appreciate, a situation that reflects the law of quantitative to qualitative change. For example, there is a rare disease, the rate of one in 100,000, if a small sample of data to see very rare, but expanded to the world’s 7 billion people, then the number is very large. In the past, technology was so backward that it was difficult to digitize the disease and focus on it. However, now that we have collected a wide variety of data cases and analyzed them in a unified way, we will soon be able to overcome many previously unimaginable scientific challenges. The list goes on and on.
It is precisely because of the many unseen values that can be mined through big data that the value of the source data itself has risen. Some companies and enterprises with a large amount of valid data have found a proven business path: selling the source data directly or through simple encapsulation. In the Internet field, social networking sites represented by Facebook, twitter, and Weibo have a large amount of user and user relationship data, and these sites are trying to commercialize and sell the source data in various ways; search companies such as Google, Yahoo! They can provide it to third parties and profit from it. In traditional industries, operators such as China Unicom (3.44,0.03,0.88%) and China Telecom (3.44,0.03,0.88%) have a large amount of underlying subscriber data, which can be simply de-privatized and then sold for profit.
The mature business model of major companies or enterprises supporting their growth by providing massive data services while compensating users with free services has stood the test of time. But for any trading of user data, there is still a need to deal with private user information and protect user privacy by deprivatizing it.
Prediction is the core of value-added services
Value-added services derived from deep mining on the basis of big data are the most imaginative business model in the big data field. What is the core of value-added services of big data? Prediction! Big data has triggered a shift in business analysis mode, from the past sample mode to the current full data mode, from the past small probability to the current large probability, so as to be able to get a more accurate prediction than before. The following more mature business models have been formed.
Personalized precision marketing. When it comes to “spam”, we are all very tired of it, because the “right” information, which is valuable in the eyes of the marketer, is sent to the “wrong” user. Through detailed analysis of a large number of user behavioral data, after in-depth mining, can be realized to the “right” user to send the “right” information. For example, large shopping malls can deeply analyze the purchase records of their members to discover the connection between users and brands. Then, when a brand’s loyal users receive the brand’s discount promotion SMS, they will not be bored, but delighted. Companies with powerful data processing technology, such as UJX, ZKJS, etc., have rich experience in data mining and precise advertisement analysis.
Decision-making guidance for business operations. Aiming at a large amount of user data, we use mature data mining technology to analyze and get various trends of enterprise operation, so as to provide strong guidance for enterprise decision-making. For example, an automobile sales company can analyze a large number of user comments on the Internet to get the features that users are most concerned about and most dissatisfied with, and then make targeted improvements to its next-generation products in order to enhance consumer satisfaction.
Overall, from the macro level, big data is the new energy of our future society; from the micro level of the enterprise, the ability to analyze and apply big data is becoming the core competitiveness of the enterprise. In-depth research and active exploration of the business model of big data is of vital significance to the future development of enterprises.
What are the characteristics of direct attached storage, storage area network, network attached storage, and IP storage network?
[Answer]: The main advantages of the DAS storage method are simple expansion of storage capacity, low investment costs, and quick results.
The main features of the NAS storage method are: attachable high-capacity storage, embedded operating system. Only provide storage services, reducing the cost of storage equipment. Particularly suitable for handling I/O requests from the network, not only fast response time, but also high data transfer rate. It can provide completely cross-platform file hybrid storage function. Its easy installation, user-friendly graphical management interface, and online expansion capability greatly facilitates the work of network administrators.
SAN is suitable for server clustering, disaster recovery and other areas of large data transmission.
IP storage network combines the advantages of both NAS and SAN: on the one hand, it adopts TCP/IP as the network protocol, which makes it easy to access the characteristics of NAS; on the other hand, it has an independent and dedicated storage network structure, and the IP-based storage network can use Ethernet technology and equipment to build a dedicated storage network.
NAS Storage Pros and Cons
NAS products are truly plug-and-play products.NAS devices generally support multiple computer platforms, and users have access to the same documents through network support protocols, thus allowing NAS devices to be used on mixed Unix/Windows NT LANs with no need for modification. The physical location of NAS devices is also flexible. They can be placed in workgroups near application servers in data centers, or they can be placed in other locations and connected to the network via physical links. Without the intervention of application servers, NAS devices allow users to access data on the network, which reduces CPU overhead and significantly improves network performance.
NAS does not address a critical issue associated with file servers, namely bandwidth consumption during backups. Unlike storage area networks (SANs), which move the backup data stream out of the LAN, NAS still uses the network for backup and recovery. one of the drawbacks of NAS is that it moves the storage transaction from a parallel SCSI connection to the network. This means that the LAN must handle storage disk requests that include backup operations in addition to the normal end-user transport stream.
Because storage data is transported over a normal data network, it is vulnerable to other traffic on the network. When there is other large data traffic on the network will seriously affect the system performance; because the storage data is transmitted through the ordinary data network, so it is easy to produce data leakage and other security problems; storage can only be accessed as a file, and can not be directly accessed like a normal file system to access the physical data block, so it will be in some cases seriously affect the system efficiency, for example, large-scale databases can not be used with the NAS.”
Please briefly describe what are the networked storage architectures and briefly explain their advantages and disadvantages?
Storage network-centric storage is a new storage architecture. It adopts a network-oriented storage architecture that separates data processing and data storage; the network storage architecture includes the essence of network and I/O, extending I/O capability to the network, especially flexible network addressing capability, long-distance data transmission capability, and efficient original performance of I/O; connecting servers and storage resources through the network eliminates the connectivity barriers between different storage devices and servers; and improves data sharing, availability, and accessibility. ; improves data sharing, availability and scalability, and manageability.
Specific parameters can be found at: https://ke..com/item/%E7%BD%91%E7%BB%9C%E5%AD%98%E5%82%A8%E4%BD%93%E7%B3%BB%E7%BB%93%E6%9E%84/10985801?fr=aladdin< /p>
Hope it can help you