CN1604528A - Cooperative starting method for network computer - Google Patents
Cooperative starting method for network computer Download PDFInfo
- Publication number
- CN1604528A CN1604528A CN 200410009789 CN200410009789A CN1604528A CN 1604528 A CN1604528 A CN 1604528A CN 200410009789 CN200410009789 CN 200410009789 CN 200410009789 A CN200410009789 A CN 200410009789A CN 1604528 A CN1604528 A CN 1604528A
- Authority
- CN
- China
- Prior art keywords
- request
- server
- file
- application program
- client
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 22
- 230000005540 biological transmission Effects 0.000 claims abstract description 26
- 238000012546 transfer Methods 0.000 claims description 11
- 230000004044 response Effects 0.000 claims description 4
- 230000007423 decrease Effects 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 description 10
- 230000015572 biosynthetic process Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005755 formation reaction Methods 0.000 description 4
- KKIMDKMETPPURN-UHFFFAOYSA-N 1-(3-(trifluoromethyl)phenyl)piperazine Chemical compound FC(F)(F)C1=CC=CC(N2CCNCC2)=C1 KKIMDKMETPPURN-UHFFFAOYSA-N 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 241000272173 Calidris Species 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000011800 void material Substances 0.000 description 1
Images
Landscapes
- Information Transfer Between Computers (AREA)
- Computer And Data Communications (AREA)
Abstract
This invention relates to a NC team startup method in network computer NC system, which comprises the following steps: the request NC sends initial procedure download requirement to the server; the server chooses the other NC as designed NC to send the files to the request NC according to the distribution information of all NC of initial application procedure; the request NC establishes a link with designed NC to accomplish the file transmission; after transmission the request NC acknowledges the server with the update information and other local stored files; the request NC judges whether all the initial procedures files are downloaded and loads the application procedure; the request NC accomplishes the startup if all the procedures are loaded.
Description
Technical field
The present invention relates to a kind ofly in network computer system, the startup method of network computer client especially has the method that the client of local computing capability utilizes the existing startup file resource of the other side to start mutually.
Technical background
The development of computer has been passed through three phases: large-scale computer stage, personal computer stage, network calculations stage substantially.Network computer (Network Computer or NC) is proposed in nineteen ninety-five by Oracle company the earliest, is representative a kind of computer form of network calculations stage.Oracle company united other four company and formulated network computer standard 1 (Network Computer RefereNCe Profile 1) jointly in May, 1996.NC has taked to simplify the way of computer hardware, and client terminal local does not have hard disk, adopts low power consumption CPU, and mainboard is provided with jumbo Nonvolatile memory devices, normally read only memory ROM or flash memory (Flash ROM).On software, traditional NC has stored Embedded operating system or terminal protocol code in the Nonvolatile memory devices of this locality.Its operation principle is by the WWW/Applet technology, application program and data on X-terminal or other-end technology (as RDP, ICA etc.) the visit remote server, and does not calculate this locality, only is responsible for receiving user's input and carries out graphical interfaces demonstration work.The calculating of tradition NC mainly occurs on the server, and is therefore very big to the dependence of server, especially when being applied to the multimedia application scene, very high to the requirement of server, so the NC number that common server can be supported simultaneously is very limited.And, because operating system is solidificated among the ROM or FlashROM of NC therefore upgrading difficulty, poor expandability.
Overweight in order to solve traditional NC to the server dependence, deficiency, software upgrading difficulty, the problem of poor expandability are supported in multimedia application.Tsing-Hua University network interconnection laboratory has proposed a kind of NC of local no operating system (referring to the applicant's relevant patent ZL03122022.3, ZL01142033.2), it has local computing capability, can greatly alleviate calculating server is relied on overweight problem, and can strengthen the NC application flexibility greatly and have very strong extended capability, can support abundant desktop application.This NC has placed NC operating system nucleus mirror image and application program on server, and disposed DHCP (Dynamic Host Configuration Protocol, DHCP) service, TFTP (Trivial File TransferProtocol, TFTP) service and long-range execution environment (Preboot Execution Enviroment, PXE) service.As Fig. 1, NC client 102 (being called for short NC or client) and NC server 103 (abbreviation server) are positioned on the computer network 101, and server provides operating system and application resource for a plurality of clients simultaneously.The start-up course of prior art NC as shown in Figure 2, client 102 is powered the start-up course that begins NC after 201, carry out the guidance code 202 in the nonvolatile storage afterwards, get in touch with server by dhcp message broadcasting, after server is received the dhcp client broadcast, reply and distribute a network address for client to client, be i.e. an IP address 203.Obtain after the network address, user end to server sends down operation system kernel mirror image and comprises the file download request 104 of the root file system mirror image of primary application program, server receives after the file download request to client transmissions operating system nucleus mirror image and root file system mirror image 105, the transmission 204 of client reception server, the NC client receives the operating system nucleus mirror image and the root file system mirror image 205 of back load server end transmission, has just finished the start-up course 206 of NC.
Though after existing NC has solved traditional NC operation well, especially calculate the too problem of reliance server under the application scenarios such as multimedia, strengthened under the run mode server to the support quantity of client.But the NC client still very depends on server in the prior art when starting: all clients all will send the file download request that NC starts to server, and download operating system nucleus mirror image and application file from server.Server not only will be handled each file download request of newly receiving, and will distribute various resources: internal memory, and virtual memory space, the I/O buffer memory, and take the CPU processing time and carry out the transmission of file.In fact, existing NC server adopts a kind of FIFO scheduling mode to carry out the processing of NC startup file download request, actual concurrent processing ability is not high, single NC server of putting for a Daepori wildcard, if receive a large amount of NC startup file download request (this multi-media classroom at NC is very common situation in using) in the short time, then server can become the performance bottleneck of startup, and the average toggle speed of NC is very slow with causing.Although and NC can share the bandwidth resources of whole network, owing to be star-like logical construction between the client and server, NC reality when starting has only utilized server to the link bandwidth resource between the client, and the utilance of network is not high.In addition, since server to the NC startup file download request received without any restriction scheme, for the server that is under the peak load, if receive a plurality of new NC startup file download requests again, server is attempted again and will be distributed above-mentioned various resources for it, will cause greater impact to server performance, cause the instability and the performance decrease of server easily, thereby robustness is not good yet, this in fact also become another server the NC number clients quantitative limitation bottleneck that can support simultaneously.
The primary application program file that NC downloads from server is that initial condition can be for the file that application program comprised of user's use after the system start-up, and it is that NC starts needed program resource except that operating system nucleus.In general, for a NC, a plurality of initial application programs are arranged, and each application program is made up of a plurality of files again, its population size size often will be much higher than the size of operating system nucleus mirror image.Therefore if server is to the processing and the transmission means of the download request of primary application program part can improve NC and start the time, then can improve the startability of NC.
Summary of the invention
Rely on overweight for server when starting at prior art NC, network utilization is not high, the problem that robustness is not good, the present invention proposes a kind of collaboration type startup method of network computer, wherein all NC and server are positioned on the computer network, store all NC on the server and started required operating system and application program, when a request NC need start, used the method that may further comprise the steps to finish startup:
Request NC is powered the back and sends broadcast on described computer network; Server is received the broadcast of request NC, and receives the response to request NC transmission, comprising the network address of distributing for request NC; Request NC receives receives the response, and sends the down operation system request to server, and server receives that this request back is request NC transmission operating system nucleus mirror image; Request NC complete operation system kernel mirror image is downloaded and is loaded, and the startup file request is intercepted, and initiates the primary application program download request to server; After receiving the primary application program download request of request NC, the distributed intelligence of primary application program file on all NC that server keeps according to this locality selected to can be other NC that request NC transmits these files as specifying NC, and informed request NC; Ask to connect between NC and the appointment NC and finish file transfer requests; Request NC sends its this locality of updating message notification server to server and has retained this document; File of every download, request NC judgement has all been downloaded as if the All Files that belongs to a primary application program and has then been loaded this application program; If all primary application programs all are loaded, then ask the startup of NC to be finished.
In addition, if request NC can't finish correct file transfer, then send and require heavy specify message to server, after server is received and is required heavy specify message, reselect another NC as specifying NC, connect between request NC and the NC that reassigns and finish file transfer requests with same file.If server can not find the NC that preserves a primary application program file, then provide the transmission of this document by server oneself.Also a buffer queue can be set on the described server, if receiving simultaneously, server enough causes the server performance primary application program download request of the quantity of decline fast, and judge do not have enough quantity NC to start, then the part request is placed buffer queue, for the request in the buffer queue, server carries out dealing with after of short duration the delaying again.
The network computer that has adopted the present invention to start method is compared with the network computer that adopts prior art startup method, the load of server obviously reduces during startup, the average toggle speed of NC is greatly improved, and network utilization also obtains bigger raising, and server has better robustness.
Description of drawings
Fig. 1 is existing NC schematic network structure
Fig. 2 is existing NC start-up course figure
Fig. 3 is a NC network diagram of the present invention
The structure chart of application file resource distributing table and client side list among Fig. 4 the present invention
Fig. 5 NC client terminal start-up of the present invention process flow diagram
Embodiment
The present invention is described in detail below with reference to diagram.
Fig. 3 has described network diagram of the present invention, and a plurality of NC clients 302 and a NC server 303 are positioned on the computer network jointly.In the start-up course, employing is carried out the transmission of operating system nucleus mirror image with the identical method of prior art, user end to server sends primary application program file download request 304 afterwards, server is handled download request, but do not start transmission to file, have the required file of requesting client but specify to client on the computer of which destination address, after client is received replying of server, initiate file transfer requests to the destination server or the client of appointment again, and finish file transfer 305 by it, therefore the transmission of file is no longer single is provided by server, and may be provided simultaneously by many computers.
As Fig. 4, the present invention has defined an application file resource distributing table 410 (AFRDT).The AFRDT that server keeps is used for dynamically NC start-up course primary application program file resource being noted down in the distribution of whole NC system (comprising NC and server); The AFRDT table that keeps on the client carries out dynamically record to the primary application program file resource of this locality; AFRDT table also is used between client-server and the client-client structure that interaction message such as the distribution, request, affirmation about the primary application program file resource provides information description.AFRDT is the list structure of a three-dimensional, the corresponding ternary relation group＜application program of its list item, file, computer 〉, wherein: the first dimension chained list 411 has write down primary application program, and wherein each node has write down the sign, title, version, state of application program, the information such as number of files 431 that are subordinate at least; A node of the second dimension chained list, the 412 corresponding first dimension chained lists, write down the information of the file that is subordinate to this application program, comprise information 432 such as sign, title, local address, state, weights, wherein weights represent to transmit the cost of this document, can adopt the parameter that is directly proportional with its size; A node in the corresponding second dimension chained lists 412 of third dimension chained list 413 has write down the information of the client of preserving this document: the sign and the IP address that comprise client.Also safeguarded a client side list 420 on the server, the index that is designated with client, write down all clients sign, title, physical network address, IP address, start information 421 such as priority level, starting state, client role, contribution margin, wherein client role can be used for operating system and/or the primary application program that server is determined client, and contribution margin is used for the client that server is selected the given transmission file.
Server started before arbitrary NC starts.On the server except preserved client all operating system nucleus mirror images that may use and initial application file resource, the table that has also kept an AFRDT structure has wherein write down the distribution of whole NC system for the primary application program file resource.When initial, do not have client terminal start-up on client also without any the primary application program file resource, so third dimension chained list be empty, this AFRDT shows in fact only to have noted down the primary application program file resource on the server.AFRDT table can be by manual configuration on the server, or is generated automatically by the configuration file that presets.Power up the corresponding document that starts and downloaded primary application program as client, then can add this client in the third dimension chained list accordingly in the AFRDT table on server.Alternatively, can be on the server according to the MAC Address of client or IP address etc., be customized parameters such as different role of different client definition and respective priority, therefore the primary application program that provides for each client also needn't be identical.
The step of startup method of the present invention before client receives the operating system nucleus mirror image adopted the way similar with prior art: the NC client is powered and begins NC start-up course (being called requesting client) after 511: carry out the guidance code 512 among the local ROM: carry out dhcp broadcast and get in touch with server on network; Server DHCP is the client distributing IP; After client obtained local IP, request was downloaded the operating system kernel mirror image from server, and the server decision is transferred to client 521 with TFTP with the LINUX kernel image afterwards to operating system that should the user; Client loads the operating system kernel mirror image of downloading from server 513, and starting the file transfer requests that TCP service may be sent by other client at any time to the port snooping that presets, requesting client is initiated a primary application program download request 514 to server afterwards.
Server receives 522 these download requests and determines its role and corresponding priority level according to the MAC Address of client, and its medium priority can be used for the file request that server judges whether this client of priority treatment, and what priority level was lower will put into buffer queue.Server is determined corresponding primary application program according to client role for it, and with information such as these application programs of AFRDT table record and corresponding files thereof, and the file request that will comprise this AFRDT table is put into suitable formation and handled.
Server maintenance two file request formations: a processing queue, a buffer queue: adopt the mode without stint of doing one's best that the file request in the processing queue is handled with the same server of prior art; But the file request that enters buffer queue then needs just can enter processing queue and handle after the time postpones δ.The δ parameter can be configured with empirical value as required, and general δ ∈ [10-60] second, or the 1-5 of the time of the single request of server average treatment after tested is doubly.
Which formation is file request put into, mainly according to server at that time cpu busy percentage P λ, memory usage M λ, the utilization rate G λ that void is deposited, performance parameter value such as output bandwidth/theoretical bandwidth P/T or carry out about the performance function value of these parameters.For example: (G λ P/T), then can decide threshold gamma ∈ (0.8,0.9) to supposition performance function Φ=Max for P λ, M λ, and Φ 〉=γ during as the processing file request then puts into buffer queue with file request; If Φ＜γ then puts into file request processing queue and handles.Alternatively, can skip the performance parameter judgement if the priority of client is low and directly put into buffer queue.
Server is handled the file request in the processing queue, can learn the needed file of requesting client according to the table of the AFRDT in the file request, then for each file in the third dimension chained list of own AFRDT table for corresponding file select preserve this document a client as (given client end) computer for this requesting client transmission this document, and be added in the third dimension chained list that the AFRDT of file request shows 523.
Server is selected the given client end in the AFRDT table of oneself, can be according to coupling first, at random, least referenced recently, or carry out based on the strategy of priority etc.Be in the consideration of fairness in addition, also can adopt the strategy based on contribution margin: in conjunction with the accumulative total contribution margin of each client in the client side list that keeps in the server, server is selected in preserving the client of this document has the minimum client of contribution margin that adds up as the given client end.If can not find any client of preserving this document, then provide transmission, in the third dimension chained list with the own AFRDT table that joins file request as intended target (hereinafter do not add differentiation for easy, all be called the given client end) by server oneself.Selection to the given client end also can be carried out in conjunction with the customization class parameters such as priority of client.Treated file request sends back to requesting client.
Requesting client is received treated file request, learns on which target from its AFRDT table and can find required file, therefore initiates file transfer requests 515 to the port that presets of given client end.After given client end 531 is received the transmission request,, then can start transmission 305 if these files are arranged oneself.Alternatively, the given client end can judge whether the self performance parameter allows just to start transmission.Client is preserved downloaded files with local ram disk (RamDisk).After a file transmission finished, it is all available that requesting client will judge whether to be subordinate to all files of certain application program, then can load this application program 516 in this way.Requesting client also will be finished message with send a transmission to server, informed in the information of message which file is requesting client downloaded at which given client end, after server is received this message, can in local AFRDT table, this requesting client be joined in the third dimension chained list of corresponding document, and in client side list, increase the accumulative total contribution margin 525 of given client end.
If requesting client is because a variety of causes can't be finished file transfer, as: can't set up TCP with the given client end and connect, or file read-write makes a mistake etc., then send and require heavy specify message 517, comprising finishing the information of the client of the file of transmission and former appointment to server.Server is reselected another given client end 524 after receiving and requiring heavy specify message.Alternatively, can simultaneously former given client end be deleted from the third dimension chained list of associated documents.Requesting client can be carried out the associated documents transmission after receiving and serving the client of thinking highly of appointment again.
If all the initialization application programs in the local AFRDT table of requesting client have all started, think that then client finished start-up course, at this moment user end to server sends a successful initiation message 518, server with the state of update request client for starting 526.
For downloading to local initial application file copy, if be in operation through client deletion or modification, then these files are not the duplicates of the document of the primary application program of the required use of other client just, therefore at this moment client also will send a file modification message to server, inform which file is no longer available, server is deleted this client after receiving this message in the third dimension tabulation of corresponding document.
For fear of when initial, many NC start simultaneously, do not exist the NC that preserves the primary application program file of sufficient amount to share the starting load of server.As a kind of improvement project, if δ in the time server receive the request that surpasses certain numerical value σ, and server judges that the startup client that does not have sufficient amount P exists, then select the request of a part (as half) is placed the buffered formation, can be during selection in conjunction with the startup priority level of requesting client.δ's is described as defined above, half of the theoretical maximum load number clients of the desirable server of σ MLC, and perhaps 2/3 of server actual max load number clients PMLC after tested, P ∈ (5%PMLC, 5%MLC).
Embodiment described above is just carried out example to the present invention, and those skilled in the art all are considered as within the scope of the present invention in the technical scheme of carrying out on the basis that does not break away from the substantive thought of the present invention obtaining after various modifications and the improvement.The suitable scope of the present invention is as the criterion with claims and determines.
Claims (3)
1, the startup method of a kind of network computer NC, wherein all NC and server are positioned on the computer network, store all NC on the server and started required operating system and application program, when a request NC need start, used the method that may further comprise the steps to finish startup:
One, request NC is powered the back and sends broadcast on described computer network;
Two, server is received the broadcast of request NC, and receives the response to request NC transmission, comprising the network address of distributing for request NC;
Three, request NC receives and receives the response, and sends the down operation system request to server, and server receives that this request back is request NC transmission operating system nucleus mirror image;
Four, request NC complete operation system kernel mirror image is downloaded and is loaded, and the startup file request is intercepted, and initiates the primary application program download request to server;
Five, receive the primary application program download request of request NC after, the distributed intelligence of primary application program file on all NC that server keeps according to this locality selected to can be other NC that request NC transmits these files as specifying NC, and informed request NC;
Six, ask to connect between NC and the appointment NC and finish file transfer requests;
Seven, request NC sends its this locality of updating message notification server to server and has retained this document;
Eight, file of every download, request NC judgement has all been downloaded as if the All Files that belongs to a primary application program and has then been loaded this application program.
Nine, if all primary application programs all are loaded, then ask the startup of NC to be finished.
2, method according to claim 1, request NC it is characterized in that: if can't finish correct file transfer in the described step 6, then send and require heavy specify message to server, after server is received and is required heavy specify message, reselect another NC as specifying NC, connect between request NC and the NC that reassigns and finish file transfer requests with same file.
3, method according to claim 2 is characterized in that: in the described step 5 and six, if can not find the NC that preserves a primary application program file, then provided the transmission of this document by server oneself.
According to any described method among the claim 1-3, it is characterized in that: have on the described server-buffer queue, if receiving simultaneously, server enough causes the server performance primary application program download request of the quantity of decline fast, and judge do not have enough quantity NC to start, then the part request is placed buffer queue, for the request in the buffer queue, server carries out dealing with after of short duration the delaying again.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410009789 CN1238990C (en) | 2004-11-12 | 2004-11-12 | Cooperative starting method for network computer |
| US11/184,466 US7467293B2 (en) | 2004-11-12 | 2005-07-19 | Method and computing system for transparence computing on the computer network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410009789 CN1238990C (en) | 2004-11-12 | 2004-11-12 | Cooperative starting method for network computer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1604528A true CN1604528A (en) | 2005-04-06 |
| CN1238990C CN1238990C (en) | 2006-01-25 |
Family
ID=34662598
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410009789 Active CN1238990C (en) | 2004-11-12 | 2004-11-12 | Cooperative starting method for network computer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1238990C (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101834888A (en) * | 2010-04-02 | 2010-09-15 | 清华大学 | Method for remotely starting transparent computing system client through wireless local area network |
| CN102148830A (en) * | 2011-03-31 | 2011-08-10 | 杭州华三通信技术有限公司 | Method for controlling flow of authentication server and authentication access device |
| CN101442432B (en) * | 2007-11-05 | 2011-11-09 | 巴比禄股份有限公司 | Network-connectable device and method for managing power thereof |
| CN102299935A (en) * | 2010-06-25 | 2011-12-28 | 纬创资通股份有限公司 | Method for improving server stability of prestart execution environment and server device |
| CN101778132B (en) * | 2010-01-22 | 2013-03-20 | 杭州华三通信技术有限公司 | Method, system and device for network installation based on load balance |
| CN103885833A (en) * | 2012-12-20 | 2014-06-25 | 中国移动通信集团公司 | Method and system for managing resources |
-
2004
- 2004-11-12 CN CN 200410009789 patent/CN1238990C/en active Active
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101442432B (en) * | 2007-11-05 | 2011-11-09 | 巴比禄股份有限公司 | Network-connectable device and method for managing power thereof |
| CN101778132B (en) * | 2010-01-22 | 2013-03-20 | 杭州华三通信技术有限公司 | Method, system and device for network installation based on load balance |
| CN101834888A (en) * | 2010-04-02 | 2010-09-15 | 清华大学 | Method for remotely starting transparent computing system client through wireless local area network |
| CN101834888B (en) * | 2010-04-02 | 2012-11-14 | 清华大学 | Method for remotely starting transparent computing system client through wireless local area network |
| CN102299935A (en) * | 2010-06-25 | 2011-12-28 | 纬创资通股份有限公司 | Method for improving server stability of prestart execution environment and server device |
| CN102299935B (en) * | 2010-06-25 | 2015-11-25 | 纬创资通股份有限公司 | Promote method and the server unit of the server stability of Pre-boot Execution Environment |
| CN102148830A (en) * | 2011-03-31 | 2011-08-10 | 杭州华三通信技术有限公司 | Method for controlling flow of authentication server and authentication access device |
| CN102148830B (en) * | 2011-03-31 | 2014-03-26 | 杭州华三通信技术有限公司 | Method for controlling flow of authentication server and authentication access device |
| US8799998B2 (en) | 2011-03-31 | 2014-08-05 | Hangzhou H3C Technologies Co., Ltd. | Methods for controlling a traffic of an authentication server |
| CN103885833A (en) * | 2012-12-20 | 2014-06-25 | 中国移动通信集团公司 | Method and system for managing resources |
| CN103885833B (en) * | 2012-12-20 | 2017-06-20 | 中国移动通信集团公司 | A kind of method for managing resource and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1238990C (en) | 2006-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200259899A1 (en) | System and method for sharing san storage | |
| EP2718837B1 (en) | Clustered file service | |
| JP3762846B2 (en) | Data processing apparatus and method for managing workload related to a group of servers | |
| US20100306280A1 (en) | Swarm-Based Synchronization Over a Network of Object Stores | |
| CN110383764B (en) | System and method for processing events using historical data in a serverless system | |
| JP5201366B2 (en) | Server function switching device, method and program, thin client system and server device | |
| US11544226B2 (en) | Metadata control in a load-balanced distributed storage system | |
| US20190147066A1 (en) | Directory structure for a distributed storage system | |
| CN100375093C (en) | Processing of multiroute processing element data | |
| EP3087722B1 (en) | Peer-to-peer network prioritizing propagation of objects through the network | |
| US10897428B2 (en) | Method, server system and computer program product for managing resources | |
| US9661042B2 (en) | Data share system, data process apparatus, and computer-readable recording medium | |
| US11922537B2 (en) | Resiliency schemes for distributed storage systems | |
| JPH06301557A (en) | Method and system for managing environment by data processing system | |
| EP1768024B1 (en) | Processing management device, computer system, distributed processing method, and computer program | |
| US20160259812A1 (en) | Method and system for accessing a distributed file system | |
| US11954362B2 (en) | Flash registry with on-disk hashing | |
| CN111404931A (en) | Remote data transmission method based on persistent memory | |
| CN1238990C (en) | Cooperative starting method for network computer | |
| JP2008107966A (en) | Computer system | |
| CN113177179B (en) | Data request connection management method, device, equipment and storage medium | |
| CN107395708A (en) | A kind of method and apparatus for handling download request | |
| US10931750B1 (en) | Selection from dedicated source volume pool for accelerated creation of block data volumes | |
| JP2004102991A (en) | File distributed transmitting and receiving system | |
| US8442939B2 (en) | File sharing method, computer system, and job scheduler |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210407 Address after: 410000 Yuelu District Yuelu street, Changsha, Hunan Province, 5 left 101 rooms of R & D headquarters of Central South University Science Park. Patentee after: HUNAN NEW CLOUDNET TECHNOLOGY Co.,Ltd. Address before: 100084 mailbox 100084-82, Beijing City Patentee before: TSINGHUA University |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220325 Address after: 410000 No. 201, building 5, R & D headquarters, Central South University Science Park, changzuo Road, Yuelu street, Yuelushan University Science and Technology City, Changsha City, Hunan Province Patentee after: Hunan youzhiduo Technology Co.,Ltd. Address before: 410000 Yuelu District Yuelu street, Changsha, Hunan Province, 5 left 101 rooms of R & D headquarters of Central South University Science Park. Patentee before: HUNAN NEW CLOUDNET TECHNOLOGY Co.,Ltd. |