Expert’s blog: How to make wise memory decision?
We face: Severe situation, and cut down the pressure of the cost. I run into the customers, will all hear such sighing with deep feeling each time.
But the problem solving method is not to just choose the cheapest data memory technology. No matter you enterprise, little company or personal domestic consumer, if you care data of you –And if you have read this article, your certain to care about your data –Then you need to consider technology and reliability in memory technology are weighed and chosen. The carrying cost is not merely the price cost of the hardware or software, it also includes operation and maintenance cost –And lose the spoiled cost of data in data.
When I begin to help the customer to understand technology weighs, the first thing that I do is to understand their requirement. Sometimes, the other side meets and looks at me blankly, then tell me that onlies want problem solving, sometimes, the other side will tell me his requirement is to obtain the cheapest memory. Can really understand there are few people of one’s own demand, and can really know less how to employ people of them.
SATA, SAS and tape
Let’s see an example chosen in different typological magnetic disc and tape drive. You may say these drivers can pass RAID independent magnetic disc redundant array Come to protect, but you also need to consider some important things. I think even if the person that calculates most does not hope either you let the data of the company in dangerous state.
The question of maximum is a technical hard error ratio. Any magnetic disc or tape drive all have a hard error ratio, this hard error ratio shows with average bits number. It may be a load too that the hard error is probably read, if this kind of mistake happens, return first false signal claim apparatus to be unable to visit. There are many kinds of reasons such as media mistake, the mistake of the magnetic head in hard Improper emergence, or media trouble. No matter what the reason is, you should consider it is that such Improper probability happen in an apparatus.
If you have one RAID-5 LUN logical unit number Take place hard error is then the the intersection of LUN and necessary rebuild, but if you happen once again hard error in rebuild, then the data will be lost. If you have RAID-6, then another hard error will not be fatal, because you have two check-up apparatuses.
The following is the hard error ratio data of magnetic disc and tape which driver manufacturers provide.
Tags: important things, error ratio, dangerous state, maintenance cost, magnetic disc, tape drive, carrying cost
Review technology of the ten innovative data restoration of 2010
Development course that the technology of data restoration has passed by the past ten years China, counting government bodies, enterprises and institutions even PC user with the skill to bring the dying back to life of this method has retrieved a large number of economic losses, play an indelible role in ensuring the data security. With the development of memory technologies such as the one of hard disk, U,etc., the technology of data restoration has got tremendous development too. In 2010, the technological industry of data restoration will research and develop numerous ” for strength very much ” Data restoration technology,if you can’t drive data restoration not to be technological, make a leap forward forward undertaking constantly. Now we choose having technology of representative ten major innovative data restoration most into it, mutually encourage with the counterpart.
The technical importance of data restoration be prominent data today severe day by day of security situation1,The cool fish of Seagate 11, 12 generations of hard disks read the technology of data restoration separately
11, 12 generations of hard disks of cool fish of Seagate were regarded as the market of hard disk of 2010 ” Protagonist ” ,Have occupied the equivalent proportion on the market of data restoration too. According to relevant investigation, in the intersection of Seagate and the intersection of trouble and plate, the intersection of magnetic head and the intersection of magnetic head and reach significant proportion trouble such as being unstable. And 11, 12 generations of hard disks of cool fish of Seagate read the technical birth of data restoration separately, not merely can shield and damage the magnetic head, read the good head data directly, the ones that have avoided the opening quotation and changed the head are heavy and complicated, save the cost, can also use technology of this data restoration through being flexible at the same time realizing which magnetic head damage is judged, disc is scratched etc..
2,Count the firmware district of hard disk and read the technology of data restoration fictitiously in the west
The fictitious data restoration technology of firmware district of hard disk of west number that introduces in October of 2010 counts the data restoration success rate of the hard disk in the west to promote greatly. Count the fictitious technology of firmware district of the hard disk through the west, with a fictitious firmware area, only rely on the firmware of own drib of the hard disk to start the initialisation of hard disk, visit the data area of hard disk directly. As everyone knows the firmware is the key data leading the hard disk to start normally, the fictitious technology in the firmware district is proposed by the global leading data restoration technology research and development mechanism efficiency source at first, it is complete to pass numerous staging, the function is strong day by day. The intersection of efficiency and source advantage put out the west count the intersection of hard disk and the intersection of firmware and the intersection of district and fictitious technology again 2010, used for solving the trouble that the ROYL series hard disk can not load the firmware.
The technology of data restoration is the important means to ensure the data security3,SAS, SCSI hard disk firmware damage the technology of data restoration
SAS hard disks and SCSI hard disks are two kinds while applying to in a large amount, data storage equipment of advanced server and top-grade work station, have the intersection of application area and wide, mulitasking, bring spacious, the intersection of CUP and the intersection of working fill and low and the intersection of heat and the intersection of person who plug and characteristic, possess higher characteristic and stability.
SAS, SCSI hard disk firmware damage one been perplexing the bottleneck of the server data restoration success rate all the time, since the global and leading organization’s efficiency source of technology research and development of data restoration put out SAS, SCSI hard disk firmware and damaged the technology of data restoration, promote the server data restoration success rate greatly, have types, have no capacity to SAS hard disk; Report ” Medium Error ” in SAS card Mistake; Sense key reports 03 mistakes; It and have type, it have capacity, way bad totally, it is getting more unable to visit to sector and reveal during measuring software such as MHDD totally on WinHex “! ” Firmware lasting the one complete solution.
Tags: hard disks, magnetic head, hard disk
Data can too ” Chill ” Data recovery method of hard disk of different class
As the hard disk break down, you may lose all data, what can you do? If data of you to worth, spend money, go, rescue, answer, the first choice is the specialized data restoration company naturally. But if you do not want to cost a large sum of money, can try to start to rescue by oneself. Dzone websites recommended one kind to make the person unimaginably queer ” The frozen law ” .
Theoretical foundation of the frozen law
The disc in the hard disk is in the high-speed rotation. When the data access or load reaches the magnetic disc, the magnetic head will be moved over into the appropriate position, extraordinary minification of distance between magnetic head and disc. According to the principle expanding with heat and contracting with cold, the frozen hard disk can further narrow the distance among disc and magnetic head, may data that can’t read smoothly read, get out again let.
Operation step:
1. Unload the hard disk from the computer.
2. Put the hard disk into the sealed bag. don’t use the low priced products
3. The hard disk already put into sealing the bag, and then put another closely knit bag. yes, you need to do it in this way, need to be absolutely dampproof
498this.style.width =498; ” twffan =” done” >4. Put the hard disk of one pair of bags in your most cold dose refrigerator.
498this.style.width =498; ” height =349 twffan =” done” >5. Frozen hard disk is more than 12 hours.
6. Will install the hard disk after refrigeration at once, answer the computer, and will begin to find the data.
7. At some time, the hard disk is still unable to read. If necessary, chill it again.
8. Might need repeating many times you,until hard disk inoperation or last all data.
Tags: break down, data access, magnetic head, speed rotation, magnetic disc, data recovery
The market of solid hard disk reproduces the battle of RAID and SLED
Over the past 20 years, when how do we set up memory system more than thing that family’s album is more important for those in planning, RAID technology has already become very deep-rooted, the majority stored insiders forgot it was once a kind of thought with revolutionary significance too. Before Patterson, Gibson and Katz published their thesis in 1988, mainframe and microcomputer have been using 14 inches of drivers exactly like Maytag washing machine all the time. Thesis of them create the intersection of RAID and the word combine 1 definition to 5 each rank. Patterson,et al. call SLED the driver ‘ Single, large size, driver costing an arm and a leg . Meanwhile, 3.5 inches of magnetic discs that the PC uses make people sympathize with very much by contrast. Think the enterprise magnetic disc whose performance of a group of small magnetic discs will exceed and cost an arm and a leg and thought defeated in the competition finally were with revolutionary significance at that time, but caused the extinction of SLED finally.
To be frank, SLED manufacturer had already had no option at that time, they could not manage to make the driver bigger, stronger and faster. Taking IBM 3380 as examples makes explanation in Patterson is in the thesis, that driver has 4 independent magnetic head positioners, can offer 200IOPs, but because technology is very complicated, cause product cost of up to 15 dollars / MB, and the consumption of single driver with capacity of 7.5GB has exceeded 6 kilowatts. Though the diameter of the driver reaches 14 inches, it is 4 magnetic heads that have offered enough spaces, but that has limited the improvement of the rotational speed of the magnetic disc. This technology has already gone to the end of the life.
The solid markets of enterprise of today make me recall that the scene when RAID competed with SLED that year arise. From EMC and IBM to Hewlett-Packard and Compellent, the majority of manufacturer has added Zeus IOPS SSDs of STEC in one’s own optic fibre passway array products, and the reason is obvious. Zeus IOPS not only can offer reading writing the data IOPS of data IOPS and 16000 of 45000, but also they have allocated the channel interface of optic fibre, so the manufacturer only needs to insert them in JBODs, then adjust the firmware of the controller admits the new apparatus.
The only problem of the flagship driver products of STEC but its price is too expensive. The capacity is probably 16,000 dollars for Zeus IOPS solid current price on market of one of 146GB, or 110 dollars / about GB, or 1 dollar / about IOPS. Because enterprise’s users will always nearly use two solid plates as one at the same time, so it is 32,000 dollars to adopt STEC solid threshold cost of one. Compared with the situation that the short distance 15,000 of the price to cheap 20-200 is transferred to the driver, it demands to put the daily data of 146GB under one’s own LUN.
Like RAID, Pillar and Equallogic have used SATA interface SLC apparatus of Intel and Samsung, writing the speed of the data is slower by about 25% than the solid products of STEC of it, but low 5% of price. If consider the cost of the slot, utilize the low priced magnetic disc of a whole set of 50 or 64GB, can provide under the same price condition for user and is twice as much as solid space and IOPS of one of two STECs.
The method meaningful to these manufacturers, like, add the intersection of STEC and the intersection of solid state and record being meaningful to EMC. EqualLogic and in try, add, get, utilize among the point-to-point the intersection of SATA RAID and array framework of controller the intersection of solid state and record Pillar. Enterprise’s users have system which support SATA, but performance is optimized on the basis of optic fibre passway driver in the passway cycle of optic fibre.
Because the diameters of the drivers of Intel and Samsung are all 2.5 inches, I think that someone will design the integrated Nehalem main board soon, LSI logic SAS/SATA RAID controller, 3U memory system of 36 Intel 25 – Es and Falconstor NSS.
STEC has SLED, Pillar and Dell / Equallogic SSD have RAID, situations will be very interesting.
Tags: magnetic disc, memory system, raid technology