The design of Tile-GX36 processor is mainly to for dealing with the server that large-scale Internet deal with. This processor can be helped and reduce energy consumption and refrigerate in the data centre the cost, can be fast at the same time to carry out the social media, search for and flow and handle the order with the multimedia.

Marketing director Bob of Tilera Company. Germany says ” Tile-GX36 processor initial clock frequency presumes as 1.2GHz, it is 24 watts to be the highly most energy-consuming. U traditional Intel compare to strong processor, Tile-GX36 the intersection of processor and each clock cycle can carry out more orders, the energy consumed is lower than traditional Intel to the strong processor too at the same time” .

Tilera processor is a central processing unit of a regular use, can be in the ordinary LINUX operating system and network data of coming up to run of application software. Whom the chip use walk abreast key quantity take less than the intersection of Tilera and 64 nuclear processor that Company introduce several years ago, but the running speed is faster than 64 nuclear processors before this. One carry eight 36 nuclear the intersection of 2U and energy consumption of server of processor in about 400 watts probably, the same as the server energy consumption which carries eight 64 nuclear Tilera processors.

“Tile-GX36 processor may run high clock frequency, more high-speed buffer memory can let each core finish more work, so we will experience higher net performance from 36 nuclear processors ” Germany says.

The analyst says, it is very difficult to do such one-to-one comparison, because the chip system framework has their own attribute. The low energy consumption from ARM and Tilera processor deals with the fast-developing cloud and is very beneficial to, occupy ruling position the intersection of X86 and processor, prove suitable to deal with such as the intersection of database and the the intersection of resource and intensive application software.

Internet trade usually run with and served by the data centre, as more and more urgent that enterprise’s users hope to compress the hope of the cost of data centre, the interest in low energy consumption server is increasing constantly too. Tilera processor already been tested on some servers, early experience’s person Hewlett-Packard and chip manufacturer Nvidia Company both choose ARM processor of the low energy consumption to make experimental server, ARM processor is mainly applied to intellectual mobile phone and dull and stereotyped computer at present. Tilera, what ARM and X86 processor is adopted is all independent order collections.

The intersection of brand and cognition degree of ARM want high a little, but Tilera processor ability further stronger in seek location by 64, Germany says. There are 32′s present ARM processors, ARM Company says they hope to introduce 64 processors of system framework of theirs in 2014, will have a far-reaching impact on server field when the time comes.

But Tilera obviously can not replace Intel in the central field of the data overnight, the processor of Tilera will go through the test of time before accepting by the market of server formally, the chief analyst Dean McCarron of Mercury research company says. It is very difficult to go to give up the server high dependability and guaranteed one but running time that X86 processor improve as to enterprise’s user.

The early attempt person will test Tilera processor in the system of such low PRI of the server of the network, because so will not influence the operation of the data centre when being busy. But the test will last several years.

” right and wrong X86 that we do not see this kind of mode, but X86 products ” McCarron expresses, he has also cited Intel X86 low energy consumption insults and moves the processor, this processor is mainly used in the experimental server which carries out tasks such as the network service,etc..

It is ” not that the market of X86 has the same demand too ” McCarron table.

These compete for system framework and all have their own advantages in energy-conserving respect like ARM and MIPS, but the overall possession cost that must consider them before comparing with system framework, the technological strategy teacher Jim McGregor of seat of honor of In-Stat Company says.

The system framework that has efficiency most will not be omnipotent either, unless enterprise’s users would like to carry on input in the software is disposed, and this part usually spends the greatest one.

McGregor shows ” it is always that some application software proves the new system framework for the energy consumption or reason of the performance “, but they, in cost an arm and a leg at software support, and speech store risk in for lock only the intersection of source and the intersection of system and user of framework ” .

“The ones that used Intel are to the strong processor, the software supports to be nearly unrestricted, have cost overallly too compared with cheap. Have AMD this kind very competitive the intersection of processor and manufacturer available at price and performance” McGregor emphasizes.

But large-scale enterprise’s user and others who dispose thousands of servers have one’s own source codes, care software support, through thrift from four to five, what major part is code recompilated very much just, all Germany say in Tilera Company.

Commonly used LINUX operating system software such as Apache network server, MySQL database and Memcached buffer the software at a high speed Can already use on Tilera processor. Tilera the intersection of system and framework can support over 2000 the intersection of LINUX and software suit, can be used for constructing the software to support too.
Share the button

For example, if one keeps the land controller with user’s data and is used as DHCP server, it is not any good idea to only use a tape to back up. The more role what the land controller acted is, the greater the danger of doing it in this way is. Have many kinds of roles’ land rare controllers, it is more common in small-scale land and medium-sized scale enterprise.

Why be with one right to take on the data server, DNS server, WINS server tape, is it a terrible idea that the land controller of roles such as DHCP server, the printing server,etc. is backed up? Mainly have some following reasons:

Your system has a dead point. If the latest tape is backed up without coming into force, you resume on having to rely on the spare data of the edition earlier to all these roles. If is unable to use tapes to resume at all, the situation will be worse.

Because the tape backs up the speed slower, can seldom back up completely. That is to say, when the disaster takes place, the last result is that some backups of the key information similar to catalogue data of the activity will be more outmoded.

It is very slow to use tapes to carry on the speed backed up too. Under the production environment, use tapes to carry on the speed resumed in an all-round way and make people unable to stand.

The and then synchronous course will consume a large amount of time.

The remaining activity catalogue server supports the operating load increasing reluctantly, performance causes anxiety.

In this text, I tell how go on, appoint roles to back up to Windows Server 2003 server you, namely the role acted according to the server backs up alone. Do it in this way, when the disaster takes place, need to resume one or more land controllers, can make the working course simpler and swifter, lighten your pressure. Meanwhile, when the tape often used is backed up to go wrong, or unable to select tapes for use to back up, still can become the straw to clutch at according to the backup that the appointed role carried on.

Point out: I propose selecting the portable medium while appointing roles to back up, but don’t use tapes. I find to any server, the value which increases a large-scale IDE apparatus can’t be appraised. I store a plurality of appointed roles to back up with it, and all valuable utility, document, brief on and helpful file.

Activity catalogue

Introduce the backup of the activity catalogue at first. Just as you understood, if you back up the systematic state data of the land controller, so movable catalogue is backed up too. The systematic state data include the catalogue database of the activity, the system starts the file, SYSVOL, registration table and some other file, role server act determine systematic the intersection of state and content of data. ,It is direct to the backup of the system state data.

Just as I mentioned in the past, about backing up the systematic state data and thing most important thing is to know that if backs up the data than the time of tombstone tombstone lifetime But also early, it is unable to resume. The tombstone time of acquiescence is 60 days. In other words, over 60 days, AD will be stored and lost. This is because tombstone lifespan determined DC to deleting for the target’s memory time. Appoint the old data on date to resume in order to prevent you from using and exceeding, the system can’t use and resume to set up over the data of one of the life. So, please check the tombstone life time on your DC to set up.

Think of resetting DNS to tremble with fear? Don’t be too anxious. If can back up correctly, even can be finished to reset DNS in the twinkling of an eye. First of all, back up the registration table correlated to DNS to set up:

Stop DNS service.

Start the editing machine of the registration table, and enter HKEY_LOCAL_MACHINE\System\ CurrentControlSet\Services\ DNS.

Some right keys on DNS file, then choose to lead out. You need to name for the file name. Our input dnsreg1 here, press and confirm.

Enter HKEY_LOCAL_MACHINE\SOFTWARE\ Microsoft\Windows NT\ CurrentVersion\DNS Server Fig. C .

Some right keys on DNS server file, then choose to lead out, here is named as dnsreg2 for the file, then press and confirm.

Save these two registration table files on your memory mediums selected.

The second step, back up and all have in the following routes on your server. The file of dns suffix: \Windows\System32\DNS keeps these file backups under the catalogue of DNS registration table file to you. though must not do it in this way, it is beneficial to put all DNS into a basket .

Finally, restart DNS service.

The remarks: If you have prefectures of the integrated activity catalogue, then the data of DNS district will form some systematic state data of DNS server.

Now, if the land controller or specialized DNS server which deposits DNS server role stops running, it is just a dish of pickles to resume DNS and set up. Please resume its elegance in the past according to the following steps. Suppose you have set up an Windows Server 2003 apparatus newly, server name, DNS suffix and IP address keep the same with apparatus breaking down.

Installation DNS service.

Bring all that you back up. The document copying of dns extension name arrives to the original position, such as \ windows\system32\dns.

Stop DNS service.

Resume two registration table files that you back up now. Need, double click two files and confirm separately only.

Restart DNS service.

Point out: Set up two land controllers at least each intra-area. In this way you could be through reinstallating the land controller resuming breaking down, this is you when it is not available to cause in the disaster, or back up the only choice while failing. Pass and reinstallate the duplication which needs the activity catalogue, so you need to have a normal land controller in the intra-area.

Take place DHCP server lose ‘ Or DHCP server role and its land controller where one stays are lost together It will be not merely a disaster, because you need heavy head to dispose the range, users keep and user’s rank. If you are paid and resumed by a disaster in the time of mid-night, this is not any good news. Fortunately back up and resume DHCP and set up the course simply. You only need one sentence to order to finish backing up:

>>dhcpback.txt

Above: >>The file that dhcpback.txt, this part, can be outputted is kept as dhcpback.txt, certainly, you can select other file names and addresses for use too.

Resume DHCP dispose, need, carry out the following, order only after the apparatus collapses completely ‘ Carry out after DHCP server service installation is finished :

It is a task often easy to be ignored too to back up WINS. Please remember, WINS will not be backed up in case of acquiescence, so as WINS database is destroyed for hour, you are unable to resume him. Fortunately realize it is quite easy too for automatic WINS to back up through disposing.

Launch WINS control panel.

Some right keys on corresponding server, then is chosen attribute.

Select the regular label page Fig. D Then in acquiescence backs up the area of route, input the spare route that you appointed, or click to have a look around and find necessarily and back up the address. The acquiescence address is % WinDir%\System32\Wins.

Choose to ” back up the database in the course of closing ” . So may guarantee no matter when stop WINS to serve and can be backed up.

Click ” confirm ” .

You can be through some right keys on relevant server too, then choose to ” back up the database ” Back up manually.

Point out: Pay special attention to, reject, keep to DNS and arrangement of DHCP in case most bad emergence of situation!

Printing server

Come to see the last role now: Printing server. I cut the subject directly here: Use the free tools Print Migrator of Microsoft.
Share the button

ARM announced a product in 2012 years and planned the street map today, showed two high-performance processors, a low consumption little controller and two figure cores at least, the code is had mostly call comes with the god, giant.

The advanced processor is key we will see ” Atlas ” Greece holds up god Atlass , ” Apollo ” Greek Sun god Apollo ,The two will all be based on the new ARMv8-A framework, support 64 to calculate, make and possibly adopt 2xnm on crafts, the applications do not all leave dull and stereotyped machine, intellectual mobile phone from the server, tabletop.

ARM shows, four manufacturers such as NVIDIA, AppliedMicro have already signed these two new central mandate, and there are two senior cooperative partners.

ARM CEO Warren East says: “We will estimate Atlas, Apollo in 2014 the putting into amount is produced, it will be the world of 2xnm craft when the time comes. ”

ARM has not revealed the concrete difference between Atlas, Apollo, but will probably make up BigLittle in a big small way as present Cortex A15, A7 ,Ordinary daily the intersection of operation and A7 close A15 at the same time, turn to A15 and, close A7 under strenuous task of calculating.

” Flycatcher” spider It is the new members of series Cortex-M, will become the processor core that ARM saves the electricity minimumly and most, suitable for fields such as the intelligence sensor, wireless sensor network node,etc..

East says: “ARM7 Thumb low the intersection of consumption and core once, Cortex-M0 one the intersection of Thumb2 and products, have lower consumptions, Flycatcher consumption low, it is unusually low. ”

ARM shows that has already found and led the cooperative partner to help them to develop this new core.

The key code of figure of future generation is ” Skrymir ” the intersection of North Europe and mythical Mil, Kerry of Sri Lankan, giant of frost, , ” Tyr ” the North European god of war mentions / Tuesday is named after him . These two cores are put in a slide and carry on the show by ARM, may imply them too the size as in pairs like A15/A7, Atlas/Apollo help each other.

East really describes as Skrymir, Tyr the high-performance core, high efficiency core separately too, but more details will be announced later by this year. He refuses to acknowledge the intimate relation of the two, but say again at the same time: ” ARM Plan to take BigLittle idea to figure field. ”

The two face various fields such as the intellectual mobile phone, mobile computing, consuming the electron, and ARM their develop, find cooperative partner right away already.

Responsible editor: King of mood

The competition makes a reservation: Operator with similar best IT professional technique
In order to really fully utilize strength of the cloud, every enterprise needs a high-performance reliable network, in order to offer the high-quality end to go to the end to serve, including users connect until holding their various IT platforms employed. So the majority of investment total of over 5 billion euro Orange Have been used in France and global network infrastructure. Get the extensive and generally acknowledged world-class telecommunication operator as network quality and performance, Orange Business Services offers remarkable service quality and security for the enterprises of various scale, make them fully utilize enormous potentiality that the cloud calculates.

The CEO Vivek Badrinath of Orange Business Services says: “Network quality is that the cloud calculates the really decisive factor of the service quality. Without totally reliable, high-performance infrastructure, the cloud is unable to fulfill its commitment. We have the ability to offer this kind of remarkable carrying to the end service quality. As IT is being merged and deepened with telecommunications constantly, the cooperation of we and more than 100 technological cooperative partners further strengthens our ability, make us offer overall cloud with strong performance service. ”

In order to provide reliable security and remarkable service quality for every enterprise, Orange and numerous authentication and standardization organizations cooperate, in order to offer cloud service and infrastructure authorized. Just because of this, first of France has obtained ISO 20000* IT and served and managed the cloud service authorized that Flexible Computing Premium has become. Orange Business Services IT in France run and also obtained ISO 9000 quality control Authorize, this company supports the centre to obtain ISO 27000 information safety certification in French Wren’s IT too. In addition, Orange Business Services has been also succeeded in finishing <>Audit the procedure, has replaced U.S.A. <>. This has further proved data centre and security, performance and reliable service continuity of network infrastructure of Orange.

It is Orange that the cloud calculates ” Conquer 2015 ” Strategic pillar of plan one of, goal its to realize, create income 500 million euro by 2015. In 2011, Orange Business Services cloud calculated the income is powerful to increase, highly agrees with this goal.

Galerie Video: The security initiated in the whole world requests the video to share
Orange Business Services is establishing the leading position of oneself in the field of cloud constantly, it announced a few days ago that would introduce first videos to share the solution in the second quarter of 2012, this solution combines the requirements for enterprise’s customer’s essential dependability and performance. The brand-new Galerie Video design of platform is suitable for all enterprise video communication resources, including enterprise video, training, on-line study, interview,etc.. This kind of concentrator can visit through Business VPN Galerie on enterprises VPN whenever and wherever possible because of Dailymotion Cloud technology. This solution serves and chains the customer’s VPN and cloud of Orange Business Services or his technological cooperative partner.

Its entry of websites do not need any special technological knowledge can carry on the customer rapidly, thus realize the sharing of the video of all enterprises, no matter how its form or size is. This service includes counting the audience for the administrator, and have function of letting users grade and comment to the video. It is based on the use of charging, the quantity based on video watched too of the monthly subscription scheme.

Unify communication and cooperate: Danone and Ond O IS has chosen Orange
Orange Business Services, with Danone and Ond separately O IS has signed the contract, in order to dispose ” business collaboration is served promptly ” Business Together as a Service ,That is to say its unified communication based on cloud serves the solution UCaaS promptly .

According to this five – year long frame agreement, Danone telephone service of the place of each running its France chartered to Orange Business Services outside. During the first stage, implementation of solution this to for Danone in the intersection of Paris and four general headquarters and 3,400 tabletop of Branch of area mainly.

This solution makes Danone group can benefit from the thing the software is served SaaS promptly Various unified communication and cooperating of the mode. Its staff can enjoy VoIP whenever and wherever possible too, dashboards, enterprise catalogue, softphone, comprehensive instant messaging and other various services of raising the efficiency shared. This kind is ” paid with the increase in demands ” Mode can last Danone after 3,400 set of initial stage tabletop if you can’t expand, serve furthering.

The industrial water circulation optimizes the specialized company Ondeo Industrial Solutions with management, for its being 60 piece different to run 300 tabletop of the place, choose Orange Business Services unified communication solution based on cloud. Branch company Suez Environnement seven countries of Europe including France of its have 650 staff.

Ondeo IS chooses ” Business collaboration is served promptly ” In order to dispose a flexible solution, it, according to using and charging, can support respectively differently to run recombination and geography in the place to move effectively. Equally important, this expanded solution can also offer the great flexibility, in order to be satisfied with the demands for the new unified communication service of staff rapidly, such as the first being open single number reach With instant messaging.

Responsible editor: King of mood

By 2013, the low energy consumption processor from ARM, these companies, will get overall application in the central field of the data, combine them with chip of the figure bring performance promoting and thrift of energy consumption greatly for enterprise layer users, the analyst says. The experience of disposing ARM processor in the server has already stepped into ripe stage, the figure processor has already been applied to some fastest supercomputers in the world too.

The analyst emphasizes, will consider the cost constantly when users are disposing the application software, so while buying the server, efficiency becomes the most key decisive factor. This year, will continue growing vigorously around cloud and high-performance server of the ultra large-scale corporation, the server will dispose the compactness reduces energy consumption and improves performance. For further energy-conservation, enterprise layer users will consider using the server which carry the low energy consumption ARM processor in the future, ARM processor is mainly applied to dull and stereotyped computer and intellectual mobile phone at present.

The research manager Jed Scaramella of IDC research company says ” users are experiencing a large amount of different technology to try to raise the efficiency ” . Enterprise’s users will pay attention to weighing each kilowatt and each square feet of expenses and server cost of per watt of performances than before.

As increasingly many enterprise’s users expand publicly-owned cloud and privately owned cloud, sell and play the function of adding fuel to the flames on the server in rise of the cloud. The enterprise users disposing the cloud around the super large-scale mode will buy a lot of pairs of way X86 servers, these new server can getting more convenient very if you can’t meddle in, expand by performance. The server under the super large-scale mode is still applied to the application software such as analysis and enterprise’s intellectual respect.

Scaramella say ” this and at will, but design for and low cost. Expanded all and how to realize around efficiency fast ” .

As the increasingly many IT administrator lays particular emphasis on density and low energy consumption demand, it replaces a choice of X86 technology in the following several years to dispose the compact server of ARM processor. Such traditional highly energy-consuming X86 server chip is a enterprise that the analyst can express the application of the low energy consumption ARM processor than from Intel to better processor or AMD white dragon’s processor Family property users are providing more efficiency for in the course that the cloud marches towards. But based on ARM processor chip want, lag behind the intersection of X86 and processor run database and the intersection of ERP and the the intersection of data and intensive software.

Some enterprises will buy the experience type server which carry ARP processor this year. Last month Hewlett-Packard announced the server which disposed Calxeda processor was designed, what this server is carried is one and four nuclear ARM processors, it is only 1.5 watts to consume energy. Great Britain Hill and Knowlton announced too the supercomputer of a Tegra 3 processor which disposed them was developing in Barcelona last month, Tegra 3 processor is a section of four nuclear ARM central processing units, can match and accelerate science and mathematical measurement with the dispersed figure processor.

“Want take place can plan street map and begin to reflect at the product from products this year. It will become reality but this does not mean gaining market share at once next year ” .

The break-through of the traditional server system structure of carrying ARM processor will be more obvious, but influence want, wait, by 2013 can reflect, add one hundred profit it consults the intersection of group and chief analyst to be red to some extent at the soonest. Ohde says. Almost all the major server manufacturers are experiencing ARM processor, Ohde adds.

Ohde says ” if Hewlett-Packard is paying close attention to ARM processor really busy, then can mean not going wrong. This still depends on how you see the development plan of ARM processor at 64 ” .

ARM processor can only support 32 to seek the location at present, there is a limitation on the error-correcting characteristic too. But the speech as to 64 ability of seeking the location, the computer can realize more large-scale memory and memory, this is very beneficial to intensive application software of data. First 64 system structure ARMv8 of microprocessor that ARM introduces at the end of October, the goal is to covering the apparatus to the advanced server field of sensor.

The idea in combination with ARM processor of the processor of the figure is very attractive too, Ohde says. The figure processor is used on a lot of world on the fastest supercomputer, much faster than performance of the traditional central processing unit too in operating some application software.

Ohde say ” the binding energy realize high density, each fast computational speed and low cost ” .

The intersection of figure and processor high-performance pet of computer, increasingly many to apply to the energy, pharmacy, financial service, in the high-performance computers of medical treatment and other trades, Ohde says.

But the analysts think that it is a challenge to carry out the transition to ARM processor from X86 processor for hardware and software. Users have already chosen to have no way of doing it against the numerous and complicated server during purchase, including storing the server, the operating load that network and software bundle together optimizes the system. It will be a subverting choice to turn to ARM processor, there are many advantages too, but also need time to dispose and suit.

Scaramella of IDC Company says ” this is a course of long journey, it is impossible to accomplish in one move ” .

The chief information officer Gerry McCartney of Purdue University responds and says, their university has carried on a large amount of input to the existing IT basic framework. The university depended on X86 server, collected writing to X86 order in a lot of codes that researchers used to a great extent.

” researchers do not hope to change the software ” McCartney expresses. The efforts that they planned to transfer from X86 system structure to other system structure are irrevocably lost a lot of years ago, a lot of ideas to the new system structure fail to realize too.

But the similar university needn’t change X86 system structure because the processor chip of Intel is having efficiency in development even more, McCartney expresses. For example, Purdue University ranks supercomputer of Carter of the 54th to carry Intel on world supercomputer’s billboard of 500 tops more than four Taiwan which they bought four years more ago It is much better that per watt of performance of the supercomputer of the processor wants.

” we will eliminate these previous four supercomputers, 1/4 that the energy consumption that uses is only consumed energy by them at present ” McCartney expresses. Collected the apparent promotion of introduction with the newest chip of every year in performance power ratio of the processor of Intel.

” have some improvements too on the processor size, we are more satisfied too in energy consumption ” McCartney expresses.
Share the button

Manager’s huge mirror seems on the discipline red day of generation of network technolohy company of old brand of Zhongguancun, there is 3,100 million passenger flow volume in all in the spring transportation this year, among them the transportation of the railway accounts for 1/10, that is to say 300 million passengers should take the railway, among them it is about 5% through the number of people that books tickets of the network, only 15 million, book tickets nervous before the Spring Festival among them, so half among them is that 8 million people booked tickets through the network. And the network books tickets the system just draws up 1 million tickets every day. The ones that cause these 8 million people not to stop in result are broken in the page, in that 1 million people get the ticket and walk from, 7 million people left will also have broken all the time. Railway book tickets network want, want, solve, visit problem totally, need, can deal with sharp point, worth 8 million some server group to hit quantity, that is to need to build 1600 servers. This quantity is not too big for a lot of IPDC data centres, Tengxun, Sina, wash treasure to be large much than this piece, if the intersection of railway and department come, plan, book tickets network can cause according to the the intersection of quantity and grade except spring transportation, other useless wasteful use of resources of most of the time, so should solve the log-in problem, also consider the question of the cost.

Then, will not buy 1600 servers on the issue that can’t solve? Certainly it is not like this.
First of all, bandwidth of booking tickets network of the railway is not enough, then a gleam of ways of line, let the thousands upon thousands of horses and soldiers take a walk, go over certainly, knock person in the front off lose line, until people bump against shed line later on oneself just, then climb and stand up to crowd inwards.

Solution: Widen the bandwidth, the railway system buys the backbone bandwidth of 200G, upgrades the server, let people log in freely, OK? We know one closing the door now rob website’s cool news of ticket, there are more than 50 servers, the network estimates that there are 15 servers at most to book tickets in the Chinese railway now, it is estimated that there are two servers too it is log-in servers, each server can accept 5000 are inserted, that is to say that there are 10000 people inside at the same time per second every second. 24*60*60*10000 =86,400,000, is close to 1 billion, disclose the data now to reveal, the clicking amount of booking tickets network is 1 billion times every day, it takes 2 servers to be newly kind under the ideal state
But the problem is, since booking tickets after the network opens, the world ranking of the network of booking tickets of Chinese railway, rise to, enter before the world 250 at once, chase one hundred mother direct, allow the intersection of Ali and father, depend on how enough mud it will be this server? More important problem is: It is a thing of the ordinary people, spends one fen more, waste too, cover oneselves and buy a electrolier lamp, the flower 1 million is justified too. Under the idea, let railway book tickets network buy 50 server, rent network bandwidth to pay 200G, you kill him, he does not do either!
Secondly, save money, too let, book tickets fast method a little? Have, that use the intersection of CDN and the intersection of network and speed technology such as Xi’an book tickets people, the server going to Xi’an after log-in, the persons who book tickets of Hebei, but the server going to Hebei, Beijing is the server going to Beijing, is not this very good? Do not turn out the person all over China, scale mountains and ford streams, step the seas and oceans far, crosses over 18 switching routers, come to the websites of booking tickets of the railway.

CDN speed technology be can rented, such as spring transportation on 40 day, 2 months at you can rent, include the mirror image servers of all parts in CDN, the bandwidth resources of all parts, bale together, rent CDN is all right!

Moreover, even if book tickets in these the server have some little trouble. Is there wood in the network card of Intel to have? There is wood with higher speed in websites to have, SSD hard disk have wood to have? One hundred degrees at present insert the server, generally adopt SSD hard disk to make the log-in server of the front desk, and put some key information here, such as active user’s database, etc., help a little of people so to come in. In addition, the front fire wall, the exchanger, between the routers, had better connect with the independent high-speed server network card, it had better be that Wan promise network card, or several network cards are placed in a server. There is acceleration that is websites again, this and CDN is not quite alike, but some measures of strengthening oneself.
In fact, these three things solve the problem method by oneself in case of not having the foreign aid, there are certain results. Buy more several yuans of the intersection of Intel SSD and hard disk, add several the intersection of Intel and high-efficient gigas of network card, several the intersection of Intel and Wan promise network card, buy several the intersection of Intel and the intersection of network and accelerator card, rent more some bandwidth, can relieve the question.
Network technique products and company mentioned inside, the rank was not successively divided, want to go who who does it say, network card Intel, network accelerator card Array, F5,etc., CDN technology offer Shanghai the intersection of network and night, the speed of Shenzhen must be opened up, there is Beijing too, servers are the treasure Germany, the tide, dawn and Lenovo’s order cycle are long. The relevant technological development of the network, or Peking University, Tsing-Hua University, Chinese Academy of Sciences. The network equipment is discipline red day of generation, this is a network company of old brand of Zhongguancun of home, with a high reputation, the service is good too, specialize in the straight conduct goods of networking products such as Intel,etc., needn’t worry buying the fake and forged thing.
Share the button

Sandy Bridge little framework make craft Cheng, its successor ” Ivy Bridge ” on the basis of 32nm Make craft Cheng, promote to more meticulous 22nm. According to Intel’s famous Tick-Tock tactics, upgrade from Sandy Bridge to Ivy Bridge, belong to the improvement of the key craft, category TOCK is improved by the little framework of Tick that that is to say ,But Sandy Bridge is TOCK category.

According to Tock Tick processor development tactics of Intel, after Nehalem, Intel processor framework is successively: Westmere Tick, 2009, 32nm ,Sandy Bridge Tock, 2010, 32nm ,Ivy Bridge Tick, 2011, 22nm And Haswell Tock, 2012, 22nm .

Sandy Bridge has adopted the name norm similar to one grade of cool Rui’s processors of tabletop to the strong processor, divided into E3, E5 and three series of E7. Among them, series E3 are the products facing the ABC of grade of forms way server, E5 series faced one pair of server platforms of way of the mainstream. E7 has adopted Westmere-EX to name.

1 2 3 4 5 Make

On the server CPU aspect, fight for the hitting out with strength of the high in the clouds cake in the face of coming from ARM framework main forces and AMD, NVIDIA, Intel the intersection of server and the intersection of processor and blueprint plan, let people pay close attention to and expect particularly 2012 year. It is reported, Intel will introduce the varied new or old framework processor in the first half of the year, will estimate that nearly there are 40 types, only Intel will introduce 7 8 nuclear processors in the first quarter ‘ Such as to strong E5-2690 , 4 fund for 6 is nuclear to strong processor, 3 4 nuclear processors and nuclear processors.

Can be seen to the products type picture of strong E5, Intel confirms the server platform processor new product blueprint for 2012 years, take the lead in releasing 3 processors such as 6 key Xeon E5-1660/E1650, 4 cores in the first season, in addition there are 7 8 key processors such as Xeon E5-2690, double-core processors such as 4 6 key processors, E5-2609,etc. 3 4 key processors such as E5-2640 and E5-2637,etc..

To the low consumption platform part of the server of small and medium-sized enterprises, will release 8 key Xeon E5-2650L and 6 key E5-2630L too. Therefore, enterprises are by ” cloud ” And expand the server and purchase and set up strong data centre demand one after another, the complete 32nm new framework products make one pair of server processor platforms of way of Intel can nearly continue keeping the absolute leading advantage.

123 makes

But, to ordinary user, what is a server? In fact the server is a high-performance computer in the network environment, it intercepts the other computers client on the network The service submitted is asked, and offer the corresponding service. For this reason, the server must undertake to serve and ensure the ability to serve.

As to enterprise, the server is various in style, come up from the framework and divide, can be divided into X86 server and not X86 servers. What daily we touched are generally all X86 servers, namely usually said PC server, it is the structure that it is based on PC organism, use Intel or other compatible X86 orders to collect the server such as IBM System x series server which deal with chip and Windows operating system, Dell PowerEdge series servers, HP Proliant series servers,etc..

In fact the server is similar to function of the PC, it is the tool helping human process information, just the localization of the two is different, the PC abbreviate as Personal Computer, PC Design to meet personal multi-function’s needs, and the server is designed to satisfy numerous users and deal with the data on it at the same time. And how do many people use the same server at the same time? This can only be interconnected through the network, come to help to achieve the goal that this use together.

Therefore, according to present state of development of network, with ” small, skillful, steady ” Get extensive application for the intersection of x86 and the intersection of framework and server of characteristic, now the intersection of server and main product of market. So the server is in purchase and network, employing the environment to be dense and integral. We can say server to born for providing service, only under the environment of enterprise’s demand IT could give play to most great value it.

1 2 3 4 5 6 Make