Opening 4 Server memory
Server memory overview
The server runs key business in the enterprise, and a downtime caused by a memory error will cause data to be permanently lost. As an electronic device, memory itself is prone to various errors. Therefore, facing the pressure of corporate facts and its own shortcomings, various manufacturers have long actively launched their own unique server memory technologies, such as HP's "online backup memory" and hot-swap mirror memory; IBM's ChipKill memory technology and hot replacement and hot-increase memory technology. As the expansion of enterprise information systems requires, the density and capacity of memory will also develop accordingly.
Server memory is also memory, and it has no substantial difference from the appearance and structure of ordinary memory we usually see in computer cities. It mainly introduces some new technologies in memory, and no conclusion can be drawn from the appearance alone. This may be worrying. If someone sells the memory stick of an ordinary PC as a server memory stick to you, it is a miscellaneous task? Generally speaking, you can rest assured that the possibility is almost zero. Because the memory on an ordinary PC is generally unavailable on the server, this also means that the server memory cannot be replaced by the memory of an ordinary PC for the sake of cheapness.
The commonly used server memory today mainly includes DDR and DDR2, and there is another type of RAMBUS memory, which is a new generation of storage products with high performance, chip-to-chip interface technology. From a technical perspective, the reason why server memory is different from ordinary memory is because of ECC, which is the abbreviation of Error Checking and Correcting. It is widely used in computer instructions in various fields. ECC is similar to Parity. However, in places where Parity can only detect errors, ECC can actually correct the vast majority of errors. Only after memory error correction can the computer's operating instructions continue to be executed. This invisibly ensures the stability and reliability of the server system. However, ECC technology can only correct single-bit memory errors. When a multi-bit error occurs, ECC memory will generate a non-maskable interrupt (NMI), and the system will automatically abort the operation.
Recommended brands: Kingston, Infineon,
Opening 5 Server Hard Drive
Server Hard Drive Overview
The server-specific hard disk is the server hard disk. We know that the server runs in a working environment with large data exchange and long working hours, so the hardware requirements are high; as a hard disk in the network data core warehouse, storing various user data and management software on it requires a safe and stable environment, so the reliability of the hard disk is very important! Nowadays, SCSI hard disks are generally used on servers because they are high-speed, stable and highly secure.
In general, the server hard disk should have high stability, high speed and the characteristics of using SCSI interface to adapt to the needs of server work.
1. High stability. Generally speaking, server hard disks mainly ensure their stability in two aspects. One is to use technology (self-monitoring, analysis and reporting technology, of course this technology is also reflected on ordinary hard drives). At the same time, hard drive manufacturers have adopted their own unique advanced technologies to ensure data security. On the other hand, there is redundant disk array (RAID) technology. In short, RAID technology stores the same data on different hard disks, so that when one of the hard disks is damaged, it can be restored from the other hard disk.
2. High speed. It is mainly achieved by adding write-after cache. Server hard drives are generally equipped with caches ranging from 2MB to 4MB, so that the average access time will be shortened and the external transmission rate and internal transmission rate will be higher. Some information mentions: SCSI hard disks that use Ultra Wide SCSI, Ultra2 Wide SCSI, Ultra160 SCSI and other standards can reach 40MB, 80MB, and 160MB respectively!
3. Adopt SCSI interface. Its full name is Small Computer System Interface, a small computer system interface. Some servers have this interface built-in on their motherboards, otherwise a SCSI interface card will be installed. Because its speed, performance and stability are better than IDE, its price is of course much more expensive and is mainly aimed at the server and workstation markets. Another main feature is that the hard disk data throughput of the SCSI interface is large and the CPU share is extremely low!
Note: SCSI hard disks are usually divided into: 68PIN, 80PIN, and by capacity: 73G, 146G, 300G
Recommended brands: IBM, Cialis
Previous page1234567Next pageRead the full text