This article refers to the address: http://
Yima Lion Network, Cheng Chengpei Shandong Cable TV CenterWith the development of digital media technology, storage technology has also been widely used in the field of digital media. In particular, the rapid changes in digital technology have made great changes in the technical means of storing digital media. While the media is experiencing explosive growth, the broadcasting and television community is very concerned about the future development of digital technology, and is also more concerned about the management, use and storage of digital media assets. This article analyzes six common digital storage technologies and their applications in digital video for reference.
SCSI technology
The development of SCSI goes through three stages. The first version of the SCSI protocol only specifies bus types, interface definitions, cable specifications, and other standards for transmission speeds of 5 MB/s. The second version made a major revision. The SCSI-2 protocol specifies a 16-bit data bandwidth. High-speed SCSI storage technology has become the mainstream of the market, and SCSI technology has firmly occupied the random storage market. The SCSI-3 protocol adds a command set that meets the requirements of a particular device protocol, making it adaptable to both traditional parallel transmission devices and to the communication needs of some of the most recent serial devices, such as Fibre Channel Protocol (FCP). Serial Storage Protocol (SSP), Serial Bus Protocol, etc.
Due to the good compatibility of SCSI technology and the large market demand, its technology is constantly being refurbished. Now it has evolved from SCSI-1 with a transfer rate of 5MB/s to 160MB/s for LVD, and recently 320 MB/s of SCSI has been put into use.
SCSI technology is widely used in non-linear editing, captioning machines and other production equipment. Early hard disk broadcast devices used this technology to build video and audio servers, but were replaced by new technologies due to their reliability and other reasons. High-reliability large-scale storage systems usually combine SCSI technology with other technologies to achieve fault self-recovery, improve security, and achieve uninterrupted work.
Storage network technology
It is a high-speed development technology in recent years, featuring high security and strong dynamic scalability. Many network storage solutions based on industry standards have been widely used. At present, the most widely used digital media field is LAN storage. In theory, the bandwidth can reach 1Gb/s, and the measured bandwidth can be around 700Mb/s. Secondly, it is Fibre Channel technology. In theory, in full duplex, the bandwidth can reach 2Gb. /s, single channel up to 1Gb/s, measured bandwidth can be around 720 Mb/s. The former is a low-cost distributed network storage solution, the latter adopts dedicated storage and gradually develops into the low-to-medium market.
Intel's Infiniband is based on the IA-64 architecture's core storage technology. The first phase is to replace the PC, the bandwidth target is 2.5Gb/s; the second phase is to reach the Cluster application, and the bandwidth target is 30Gb/s. Whether the goal is ambitious, whether it can be accepted by the market, and whether the technical bottleneck can be broken, people will wait and see.
Storage network technology has developed rapidly in the video field in recent years. Whether it is from management, production or broadcast, it is widely used. However, under the current technical conditions, the formation of a large-scale TV station's production and broadcasting integrated network, the central storage and unified management of the entire Taiwan media assets, there are still many technical difficulties to overcome. In particular, the bandwidth problem of storage network technology is the biggest obstacle.
RAID technology
RAID is a redundant array of multiple inexpensive disks. Although RAID contains multiple disks, it appears as a separate large storage device under the operating system. RAID technology is divided into several different levels to provide different speeds, security, and price/performance ratios.
RAID0 is the simplest form. RAID0 can connect multiple hard drives together to form a larger storage device. However, because RAID0 has no redundancy or error repair capability, its security is greatly reduced. Therefore, it is not sensible for general applications to configure more than four hard disks in RAID0. If any of these disks fails, the entire system will be destroyed and cannot be used anymore. In the early days of the country, some video and audio servers used RAID0 technology, and few of them were able to be used for a long time and safely.
RAID1 and RAID0 are completely different, and their technical focus is all on how to ensure system reliability and data repairability without compromising performance. RAID 1 is also known as disk mirroring, and each disk has a corresponding mirror disk. RAID1 is the most expensive implementation of all RAID levels. However, people choose RAID1 to store critical and important data.
RAID3 uses a dedicated disk to store all the parity data, and creates a read and write operation of the striped data in the remaining disks. RAID3 not only provides fault tolerance like RAID1, but the overall overhead drops from 50% of RAID1 to 25% (RAID3+1). As the number of disks used increases, the additional cost overhead will become smaller and smaller. In different situations, the complexity of RAID3 read and write operations is also different. The simplest case is to read data from a good RAID3 system. At this time, only need to find the corresponding data block in the data storage disk for reading operation, without adding extra system overhead. When writing data to RAID3, the situation becomes more complicated. Even if we just write a block of data to a disk, we must calculate the checksum of all the blocks in the same strip as the block and rewrite the new value into the check block. From this we can see that a write operation actually includes four steps: data read (read associated data block in the strip), check value calculation, data block write and check block write. The system overhead is greatly increased. We can simplify the RAID system by setting the size of the strip appropriately. If the length of a write operation is exactly equal to the size of a full stripe (full stripe write), then we do not have to read the associated block in the strip to calculate the checksum. We only need to calculate the check value of the entire band, and then directly write the data and check information to the data disk and the check disk. So far, what we are talking about is reading and writing data under normal operating conditions. Below, let's look at the operation of the RAID system in degraded mode when the hard disk fails.
Although RAID3 is fault tolerant, system performance is affected. When a disk fails, all data on that disk must be re-established using the checksum information. If we read the data block from a good disk, there will be no changes. However, if the data block we want to read is located on the damaged disk, we must simultaneously read all other data blocks in the same band and reconstruct the lost data according to the check value. When we replace the damaged disk, the system must rebuild the data in the bad disk for one block of data. The entire process involves reading the band, calculating the missing data blocks, and writing new blocks to the new disk, all automatically in the background. Rebuilding activities are best done while the RAID system is idle, otherwise the performance of the entire system will be severely affected.
Unlike RAID3, RAID5 distributes the parity data evenly to each disk. Each hard disk is stripe-divided separately, and the same stripe area is parity-checked (exclusive OR), so that any pair can be ensured. The read and write operations performed by the parity block are balanced across all RAID disks. As a result, RAID 5 has good random read performance because data within a specified transport block size range only requires access to a single data drive and overcomes the limitations of RAID3 single redundant disks.
The main disadvantage of RAID5 is that it reduces the write function because it is written to the disk one bit or one byte. After processing, the bit or byte of data block 1 is written on data block 1, the bit of data block 2 or Bytes are written on data block 2, ... Therefore, there are more links to be processed when writing data, which reduces the random write function.
The technology of building RAID with IDE hard disk is a new technical direction. Due to the scalability of IDE devices and the technical limitations of IDE devices that support hot plugging, RAID applications for IDE devices are not yet widely available.
In the broadcasting industry, the most used RAID technology is the video and audio server and the non-linear editing hard disk tower. Other storage devices are also widely used, but not as compelling as the above devices.
SAN technology
SAN is the product of the era of storage technology entering the network. On the one hand, it can provide rich, fast and simple storage resources for network application systems; on the other hand, it can centrally and uniformly manage online storage resources, and become the ideal storage management and application mode. It can be used as a structure for TV station business management or as a networked framework for video and audio broadcast servers.
NAS technology
NAS is one of the fastest growing data storage devices. In a typical network architecture, data becomes the center of the network, and NAS devices are directly connected to the network. It has the following characteristics: (1) The NAS device exists as a separate file server. Most of the data for all devices in the network is stored on the NAS device. (2) It is very convenient to connect the NAS device to the network. If you set up a simple IP address, you can use the NAS device plug and play. (3) The convenience of using NAS equipment can greatly reduce the management and maintenance costs of the equipment. (4) NAS devices can support different operating system platforms. At the same time, it provides RAID hard disk, redundant power supply and fan, redundant controller, which can guarantee 7×24 hours of operation.
This technology is used in the field of digital video for central online storage, network hard disk servers and network nonlinear editing.
Data stream tape technology
Streaming tape technology is an ancient technology. With the continuous development and updating of technology, its capacity, read/write speed and reliability have been rapidly improved, and its application in the field of broadcasting and television has also attracted attention. There are three types of commonly used tape storage technologies:
1.LTO technology
LTO, the Linear Tape Open Protocol, was jointly developed by HP, IBM and Seagate at the end of 1997. It is an open technology, and the three manufacturers open production licenses to storage media and tape drive manufacturers to make products compatible with different manufacturers. Openness brings more innovation, is compatible with existing equipment, reduces costs and prices, and benefits users. LTO combines linear multi-channel, bi-directional tape formats, hardware data compression, optimized track storage and efficient error correction to dramatically improve tape performance. LTO currently supports Ultrium (High Speed ​​Open Tape Format) and Accelis (Quick Access Open Tape Format). The Ultrium format features high reliability and high capacity. In particular, it can be operated separately or in an automatic environment. Accelis focuses on fast data storage. It has dual-track tape storage in the tape cartridge for faster reading and writing. Both formats use the same head, media track faces, channels, and service technologies, and share a lot of code. In contrast to the two formats, most users emphasize storage capacity, making Ultrium technology even more compelling. The domestic radio and television industry began to use this technology for media asset storage in 2002, which is a technical trend worthy of attention.
2.DAT technology
DAT (Digital Audio Tape Technology) was first developed by HP and Sony. It uses spiral scanning technology, which was mainly used for digital audio storage in the early days, and was later improved for information storage, and there are indications that the advantages of DAT will continue to be maintained. The reason why DAT technology is very popular is because of its high cost performance and high reliability. In addition, the technology is used all over the world, so the continuous supply and good after-sales service of the product are available worldwide.
3.DLT (Digital Linear Tape) Technology
The technology was first developed by DEC in 1985 and is mainly used in VAX machines. At that time, it was high performance and high price, and it was only used in a few fields. After improvement, it has become a hot technology in the storage field. At present, the tape drive capacity is 10 to 35 GB, and the hardware compression technology can double the capacity. But DLT technology also has certain disadvantages. The price of the drive and magnetic media is high, and the bandwidth between the main system and the network is narrow. The non-standard exterior design limits the interior. But the technology can still be seen as a promising product in the future. Currently only used by a small number of users who need high performance backup.
It is particularly worth mentioning that the above various data stream tape drives are said to have hardware compression technology, which can double or double the data losslessly. However, in the field of radio and television digital media, most digital media uses MPEG-1, MPEG-2 and MPEG-4 compression. When these data are saved to the tape drive, the data can no longer be compressed. If you use the hardware compression of the tape drive again, you can not only compress the data, but also increase the capacity, which is the result of the experiment. When considering the capacity of a tape library, if you store MPEG or JPEG files, you cannot design capacity by hardware compression.
In addition, optical storage technology, Cluster storage, IP storage and object-oriented network database storage technology are also booming and worthy of attention, limited by the space here.
In fact, the above technologies do not exist in isolation, but the application of these technologies to construct systems. For example, the storage array servers of GVG and Pinnacle AV servers use RAID3 technology, storage array server primary and backup images, and AV servers and storage array servers adopt FC NAS. The storage array of SeaChange's server adopts RAID5 technology. The server uses Cluster connection and RAID2 fault self-recovery mode, and the external interface is interconnected by Local Ethernet.
When selecting and designing a system, consider the scalability of the storage and system. For example, a networked nonlinear editing system, if the hard disk storage array uses single-channel FC transmission, the theoretical bandwidth is 1000 Mb/s, and the actual transmission effective bandwidth is about 70% of the nominal bandwidth. On such a system, 8 or 8 More than one fine editing station is prone to wait or crash when working at the same time (when the fault-tolerant function is not perfect). If you order a single-channel FC hard disk storage array design, and later expand more than 5 fine-tuning stations without expanding the structure of the hard disk storage array, it is normal to have problems.
Technology is developing, and multiple storage technologies compete with each other, promote each other and integrate, and survive. There is no absolute advanced lead, only a relative lead. Especially when designing a digital media application system, it is necessary to determine the storage technology and structure type according to the object, investment, and technical requirements. The lack of foresight and excessive pursuit of low price will cause unnecessary waste.
Imported PMMA material upper cover, which has extremely good transparency and strength.
The tank diameter has options of 50mm and 60mm, and the length provides the options of 60m, 130mm, and 190mm.
The POM box provides the options of black and white, the metal material box cover provides the four options of red, green, blue and silver.
The T-virus series Water Tank has a spiral-shaped double pipelines with cooler shape.
The water tank connectors are standard G1/G screw threads with good compatibility.
By matching RGB bar lamps, it could realize multi-color light effect (bar lamp supports the Syscooling lamp control system).
Mature sealing process and advanced sealing material to ensure the using safety.
It is suggested to use the special heat conducting liquid of Syscooling.
Transparent Water Tank,Plastic Water Tank,Water Tank,Computer Water Tank
Dongyuan Syscooling Technology Co., Ltd. , http://www.syscooling.com