CN1294509C - Cluster computers possessing distributed system for balancing loads - Google Patents
Cluster computers possessing distributed system for balancing loads Download PDFInfo
- Publication number
- CN1294509C CN1294509C CNB021320276A CN02132027A CN1294509C CN 1294509 C CN1294509 C CN 1294509C CN B021320276 A CNB021320276 A CN B021320276A CN 02132027 A CN02132027 A CN 02132027A CN 1294509 C CN1294509 C CN 1294509C
- Authority
- CN
- China
- Prior art keywords
- node
- computer
- cluster computer
- server
- cluster
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Hardware Redundancy (AREA)
- Computer And Data Communications (AREA)
Abstract
The present invention relates to a cluster computer, particularly to a cluster computer with a distributed load balancing system. When the cluster computer makes a calculation, the calculation load of each node can be balanced by the control of the distributed load balancing system so as to enable all the nodes to access the same storage system, and storage equipment included in each node is reduced to lower the threshold for the introduction of the cluster computer. Meanwhile, when a failure in a front-end server of the cluster computer occurs, the distributed load balancing system can control each piece of equipment and each node, and the calculation operation of each node is continued, and is not influenced; besides, calculation data is properly and safely protected, so the storage or the backup of the data is more convenient; moreover, the processing speed and the work efficiency of the cluster computer are enhanced.
Description
Technical field
The invention relates to a kind of cluster computer; be particularly to a kind of cluster computer with distributed system for balancing loads; allow cluster computer when carrying out computing (as indivisible computing); the control that sees through distributed system for balancing loads can make the common same stocking system of access of each node (Node); but and the various data in safekeeping cluster computer when running, and the data processing speed that promotes node.
Background technology
Along with the progress of semiconductor fabrication and integrated circuit (IC) design, personal computer extensively has been used in aspects such as individual, family, academic research, commerce and industry.On the other hand, flourish along with Internet, bulk information constantly flows in Internet, particularly in the field of ecommerce and academic research, information processing and flow frequent especially and heavy, so need processing power powerful and relevant device such as high-order server and equipment that have a high-reliability to carry out stable support and running.
Yet its price of high-order server that generally has powerful processing power is but very expensive, and does not meet economic benefit of enterprises.Because the processor that at present general personal computer is used has had to a certain degree processing power and reliability, and by using superfast network on-line (as network with Gigabit transfer rate, or even the network on-line of application fiber channel and optical network unit foundation), data transfer rate between computing machine and the computing machine significantly improves, so can be by technology such as parallel processing and Distributed Calculation, many personal computers are construed as a cluster computer with the network type of attachment, provide high-level efficiency calculating and processing to give each field.
Please refer to Fig. 1, its illustrate is the synoptic diagram of known cluster computer.Known cluster computer 10 mainly is made up of a plurality of nodes 30 and at least one principal computer 70, and each equipment is to see through network access device (as interchanger etc.) 60 to be connected to each other to constitute a LAN 20, wherein principal computer 70 has the load balance of Task Distribution, Control Node 30, and function such as all runnings in the monitoring cluster computer 10, with in being controlled in case of necessity.
Please refer to Fig. 2, its illustrate is the equipment connection diagram according to the intranodal of known cluster computer shown in Figure 1.In the known cluster computer 10 of Fig. 1, each node 30 can be connected to each other in the mode of shared bus 100 by one or more processors 80 and memory module 90 to be formed.
And each node 30 of known cluster computer 10 all has exclusive node storage facilities 40 or node stocking system etc., as disk array (RAID), network storage equipment (Network AttachedStorage; NAS) and store LAN (Storage Area Network; SAN) etc., the data so that storage node 30 is handled perhaps are used as data backup.Generally speaking, if node 30 is when being connected a plurality of processors 80 and constituted by shared bus 100, node storage facilities 40 mostly is the shared storage facilities that is connected to shared bus 100, for a plurality of processor 80 accesses simultaneously.
Generally speaking, after node 30 was finished the subtask that principal computer 70 pays, principal computer 70 will be paid another subtask and carry out for node 30, till this computing (as indivisible computing) is finished according to the principle of balancing dynamic load.When this computing was finished, principal computer 70 will be according to the design of application program, extracted the result of the processor active task that nodes 30 carry out from the node storage facilities 40 of each node 30, and it is integrated the unified result of computing that is.Perhaps when each node 30 is finished its subtask of carrying out, earlier the result is sent to principal computer 70, principal computer 70 wait to obtain all subtasks as a result the time, the result of all subtasks is integrated unification, to obtain the result of processor active task (as indivisible computing).
Each node 30 all can carry out data access to node storage facilities 40 when handling the subtask, therefore, the work efficiency of each node 30 and node storage facilities 40 are not without connection.For example, the running speed of node storage facilities 40 and frequency range between processor 80 is connected, node storage facilities 40 and the workbench of cluster computer 10 and application program all can influence the work efficiency of node to optimization of node storage facilities 40 etc.
In addition, when cluster computer 10 carries out computing (as indivisible computing), each node 30 is when carrying out processor active task, need while access node storage facilities 40 and processing communication from principal computer 70, for example 70 pairs of nodes of principal computer 30 carry out balancing dynamic load, principal computer 70 inquires that 70 pairs of nodes 30 of progress, principal computer of nodes 30 processor active tasks require the result of processor active task etc., make node 30 fail to be absorbed in the processing processor active task, the work efficiency of node 30 thereby decline to a great extent.
Aspect node 30 foundation or safeguarding, because each node 30 all needs exclusive node storage facilities 40, so when increasing node 30 newly, cost certainly will rise because of the node storage facilities 40 of costliness, and, numerous node storage facilitiess 40 are that correspondence depends on each node 30 respectively in the cluster computer 10, capacity based on storage facilities is big more, the price principle that the cost of each storage unit is cheap more, the cost of distributed node storage facilities 40 is higher, the convenience when also having reduced simultaneously the security of storage data and Backup Data.On the other hand, for the work efficiency that improves cluster computer 10 and increase its stability, the specification of each node storage facilities 40 must be unified as far as possible, thereby has limited the elasticity of buying or replacing node storage facilities 40, and the cost that has cluster computer 10 just is difficult to reduce.
Aspect the data preservation, because principal computer 70 is to be responsible for functions such as control task distribution and load balance, when decommissioning as if principal computer 70 catastrophic failures in the cluster computer 10, each node 30 just can't be guaranteed normally inerrably to operate, data that node 30 computings simultaneously are required or the data that produced are Be Controlled not also, also may be affected and damage and be stored in data in each node storage facilities 40, visible known cluster computer 10 is failing to provide stable fiduciary level aspect the data preservation.
Along with the development of kownledge economy, the application demand that enterprise handles for mass data is urgent day by day, and demands such as the processing of data, storage, safety and management also present explosive the growth, and the application of cluster computer is with increasingly extensive.Yet that the enforcement of conventional cluster computing machine and use but fails to provide simultaneously is low-cost, the computing of high-reliability and high workload efficient.Therefore, how to set up can save enterprise cost, strengthen expansion capacity, the cluster computer of tool high stability, high security, be era of knowledge-driven economy enterprise's key of winning victory of being able in cut-throat competition, show one's talent.
Summary of the invention
In above-mentioned background of invention, development along with kownledge economy, the application demand that enterprise and academic research etc. are handled for mass data is urgent day by day, and demands such as the processing of data, storage, safety and management also present explosive the growth, yet its price of high-order server farm is but very expensive.But progress along with semiconductor manufacturing, integrated circuit (IC) design and communication network technology, general personal computer has possessed reliable processing power, the speed of network service is fast lifting also, therefore utilize general computing machine to see through network and connect the cluster computer that construction forms, ability to work near the high-order server farm can be provided, but foundation, maintenance or expansion cost are all cheap than high-order server farm, in having provided another embodiment outside the server farm.In cluster computer, each node all has the usefulness that exclusive node storage facilities or node storage network are done the data access, but when increasing node newly, cost certainly will rise because of the node storage facilities of costliness, and the numerous node storage facilitiess that disperse in the cluster computer, cost greatly increases when also causing Maintenance Point.
In addition, for the work efficiency that improves cluster computer and increase its stability, the specification of each node storage facilities must be unified as far as possible, has limited buying or has changed the elasticity of node storage facilities, and the cost that has cluster computer just is difficult to further reduction.
Moreover, if control task is distributed and the principal computer of functions such as load balance hinders and when decommissioning for some reason in the known cluster computer, the running of each node and monitoring, and internodal load balance just is affected, the data that are stored in each node storage facilities also may be damaged, valuable data thereby loss.
Fundamental purpose of the present invention is for providing a kind of cluster computer, particularly relevant for a kind of cluster computer with distributed system for balancing loads, allow cluster computer when carrying out computing (as indivisible computing), the same stocking system of the common access of each node, reduce the storage facilities that each node comprises, to reduce the threshold that imports cluster computer.Simultaneously, also can promote the processing speed of cluster computer and degree of operating reliably, and can increase and store or security and convenience during Backup Data.
According to the purpose of the above, the invention provides a kind of cluster computer with distributed system for balancing loads, this cluster computer with distributed system for balancing loads comprises at least: the network connection device (as interchanger) with a plurality of connectivity ports; Be connected to the stocking system of network connection device; Several nodes, these nodes are used for carrying out a computing (as an indivisible computing) jointly, and each node is to be connected to network connection device, when carrying out computing stocking system is carried out access; At least one front-end server, the one end is connected to Internet, and the other end then is connected to network connection device; And at least one load balance server, one of this at least one load balance server end is to be connected to network connection device, when the front end server is appointed processor active task when giving this at least one load balance server, this at least one load balance server then carries out processor active task distribution and load balance control to these nodes, with the subtask mean allocation of processor active task to each node, and the running of control stocking system, to optimize the running of the cluster computer with distributed system for balancing loads provided by the present invention.
In addition, the present invention's the cluster computer with distributed system for balancing loads more includes control desk, in order to the running of monitor network connection device, stocking system, numerous node and at least one computer server.
And the stocking system in the present invention's the cluster computer with distributed system for balancing loads can constitute by storing LAN or network storage equipment.In addition, each node can be made of several processors and at least one memory module connected mode with shared bus.
Utilize and use the centralized stocking system of distributed system for balancing loads control and the operating load of node in the cluster computer, the processing power in the time of can promoting node work is accelerated arithmetic speed.Simultaneously, when front end server running is broken down and when stopping, utilize the control of distributed system for balancing loads to each equipment and node, the present invention's the computing that cluster computer carried out still can continue to carry out without a break, and because each node is that same centralized stocking system is carried out data access, so the operational data of cluster computer is under lock and key in centralized stocking system, not impaired because of the change of external environment, so the management of data, store and backup safety and conveniently more just.In addition, by utilizing centralized stocking system, the cost when setting up cluster computer or expanding node is significantly reduced.
Description of drawings
Preferred embodiment of the present invention will be done more detailed elaboration in conjunction with the accompanying drawings in following comment, wherein:
What Fig. 1 showed is the synoptic diagram of known cluster computer;
Fig. 2 shows is equipment connection diagram according to the intranodal of known cluster computer shown in Figure 1;
What Fig. 3 showed is the connection diagram of one of the present invention embodiment;
What Fig. 4 showed is the connection diagram of one of the present invention preferred embodiment; And
What Fig. 5 A to Fig. 5 B showed is according to the synoptic diagram of entreating the embodiment of stocking system among Fig. 4.
Embodiment
Please refer to Fig. 3, what it was shown is the connection diagram of one of the present invention embodiment.As shown in Figure 3, cluster computer 300 provided by the present invention, it is the node 320 that is connected to LAN 310 by a plurality of, front-end server 370 that connects Internet 380 and at least one have distributed system for balancing loads and form as the server 340 of work such as load balance, and have more a central stocking system 330 that is connected to LAN 310, wherein each node 320 can be connected to each other in the mode of shared bus by one or more processors and memory module and form, and does not comprise node storage facilities (or node stocking system) in each node 320.
When cluster computer 300 runnings, front-end server 370 is used as processor active task (as indivisible computing) server 340 of work such as load balance, then by server 340 with this processor active task mean allocation to each node 320, by the subtask of numerous nodes 320 these processor active tasks of execution, to reach purpose to load balance.Because in the present invention's the cluster computer 300; be to utilize enforcements such as different servers and the network equipment to be connected the front-end server 370 of Internet 380 respectively and as the server 340 of load balance; and front-end server 370 has the function of fire wall, so can protect the running of cluster computer 300 inside.
Simultaneously, through the running platform of suitably revising and adjust cluster computer 300 and application program etc., for example, because node 320 and central stocking system 330 all have medium access control address (Media AccessControl Address; MAC Address), thus can this MAC Address as the reference of streams data, when node 320 carries out the subtask of processor active task, desired data is stored to central stocking system 330, perhaps read desired data from central stocking system 330.
In addition,, need not be responsible for front-end server 370 work of load balance because just being responsible for processor active task is used as the server 340 of load balance, so attack of being failed to understand when front end server 370 or fault and when decommissioning, have the still sustainable running of server 340 of distributed system for balancing loads and administer each equipment and the node of cluster computer 300 inside, continue node 320 carried out work such as load balance, so the running of node 320 is unaffected.And, because the present invention's cluster computer 300 is to adopt central stocking system 330, so control via server 340 and application program, required data or deposit the operation result that is produced in and be not subjected to ectocine when node 320 can continue that central stocking system 330 taken computing, the data that are stored in central stocking system 330 also can thereby not run off, and important data are just properly saved from damage.
In addition, front-end server 370 and server 340 can be directly after 330 pairs of node 320 computings of central stocking system the result of gained take domination, and do not need to upload operation result, and need the communication node just can obtain the operation result of node as the principal computer wait node of known cluster computer.Therefore, front-end server 370 and server 340 do not need usually to inquire the progress of node 320 processor active tasks, also do not need node 320 is required the result of processor active task, node 320 just can utilize whole resources to be absorbed in the processing processor active task, so the work efficiency of node 320 is promoted, the treatment efficiency of cluster computer 300 also significantly increases thereupon.
Please refer to Fig. 4, its demonstration be the connection diagram of one of the present invention preferred embodiment.In Fig. 4, the present invention's cluster computer 300, be to utilize network access device (as interchanger etc.) 350 to connect a plurality of nodes 320 simultaneously, two (or more than two) are as the server 360 of work such as load balance, two (or more than two) connect the front-end server 370 and the central stocking system 330 of Internet 380, to constitute a LAN 390, wherein each node 320 can be connected to each other in the mode of shared bus by one or more processors and memory module and form, but each intranodal does not comprise node storage facilities (or node stocking system).
And, more comprise control desk 400 in the cluster computer 300, utilize modes such as privately-owned network is online, directly link and communicate by letter with front-end server 370, server 360, central stocking system 330 and the numerous nodes 320 etc. that are used as work such as load balance, grasp the function situation of each equipment in the current cluster computing machine 300, so that managerial personnel monitor at any time.
Because the cluster computer 300 of Fig. 4 has the channel that connects toward Internet 380, so front-end server 370 generally all has firewall functionality, and the reason that adopts the front-end server 370 of two (or more than two) is used for the work of on average sharing, and when wherein a front-end server 370 is failed normal operation, another front-end server 370 can be taken at once, and the channel of avoiding connecting Internet 380 is paralysed fully.In like manner, the server 360 of cluster computer 300 employings two (or more than two) is supported mutually and is standby, degree of stability when increasing the cluster computer running, and another can be supported to control each node 320 and continues to operate when arbitrary station server 360 faults.
In addition, please also refer to Fig. 4, Fig. 5 A and Fig. 5 B, Fig. 5 A and Fig. 5 B are shown as the synoptic diagram according to the embodiment of centre stocking system among Fig. 4 respectively.Centre stocking system 330 is realized to store LAN (SAN) 500 among shown in Fig. 5 A, storage data is sent to the opertaing device 510 that stores LAN 500 through connection line 410 automatic network connection devices 350, and opertaing device 510 is stored to suitable storage facilities 520 with storage data then.
Use the present invention's cluster computer 300, except the work efficiency of node 320 is promoted, owing to do not comprise node storage facilities (or node stocking system) in the node 320 of cluster computer 300, so cost is reduced when expanding node 320, but can not reduce the work efficiency of node 320, and because of the centralized central stocking system 330 of cluster computer 300 employings, but not be distributed node storage facilities in the known cluster computer, capacity based on storage facilities is big more, the price principle that the cost of each storage unit is cheap more, can learn that the storage cost of using central stocking system 330 is comparatively cheap, and the not good problem of utilization rate of storage facilities can improve node 320 and use exclusive node storage facilities (or node storage network) time.
In addition, owing to used centralized central stocking system 330, the composition of cluster computer 300 is simplified, and the composition of its application program and job platform is also simplified, and just can improve the speed of execution.Factor is increased the convenience of storage management personnel when safeguarding and back up greatly according to managing concentratedly and storing, and has also promoted the unified and security of integration of data simultaneously.
The present invention's advantage is for providing a kind of cluster computer with distributed system for balancing loads, because the present invention's cluster computer is to be appointed and external online front-end server by responsible processor active task, be responsible for the load balance server of processor active task distribution and load balance, node and central stocking system are formed, so when externally online front-end server fault, do not influence load balance server in the cluster computer, the running of equipment such as node and central stocking system, unaffected so the computing that cluster computer is carried out can continue to carry out, be stored in central stocking system simultaneously and also can not run off because of the front-end server fault.And, because each node is the same central stocking system of access all, but node just pooling of resources is handled computing, so the processing power of node is improved, arithmetic speed also can be accelerated, make that simultaneously the management of data is more convenient with backup, and reduce the cost when setting up cluster computer or expanding node.So the present invention's cluster computer provides the computing of low cost, high-reliability and high workload efficient.
In addition, since the present invention's cluster computer with distributed system for balancing loads have set up and user (Customize) easily, lowly import cost, high expansion elasticity but lowly expand characteristics such as cost, high reliability and degree of stability, so be particularly suitable for the usefulness that enterprise, academic research and engineering application etc. need the field of high-level efficiency calculating.
As the personnel that are familiar with this technology understand, and the above only is the present invention's preferred embodiment, are not in order to limit the present invention's claim; All other do not break away from following equivalence of finishing of disclosed spirit and changes or modification, all should be included in the following claim.
Claims (7)
1, a kind of cluster computer with distributed system for balancing loads comprises at least:
One network access device, this network access device has a plurality of connectivity ports;
One stocking system, this stocking system has expandability, and be correspondence be connected to described a plurality of connectivity port one of them, and this stocking system is one of to store in a LAN and the network storage equipment person with one to implement;
A plurality of front-end servers, one of this each front-end server end is to be connected to an Internet; The other end correspondence of this each front-end server is connected to one of described connectivity port, and wherein these a plurality of front-end servers are supported mutually and be standby;
A plurality of nodes, described a plurality of nodes are used to carry out a plurality of subtasks of a processor active task, and this each node is the corresponding connectivity port that is connected to these a plurality of connectivity ports, when carrying out this a plurality of subtask this stocking system is carried out an access step; And
A plurality of computer servers, one of this each computer server end is corresponding to be connected to again one of described a plurality of connectivity ports, and these a plurality of front-end servers are used for this processor active task is appointed to these a plurality of computer servers, and these a plurality of computer servers are used for these a plurality of nodes are carried out processor active task distribution and load balance control, and wherein these a plurality of computer servers are supported mutually and be standby.
2, the cluster computer with distributed system for balancing loads as claimed in claim 1, wherein above-mentioned network access device comprises an interchanger.
3, it is with this a plurality of subtasks mean allocation of this processor active task these a plurality of nodes extremely that the cluster computer with distributed system for balancing loads as claimed in claim 1, wherein above-mentioned processor active task distribute.
4, the cluster computer with distributed system for balancing loads as claimed in claim 1, wherein above-mentioned cluster computer more comprises a control desk, in order to monitor the running of this network access device, this stocking system, these a plurality of nodes and these a plurality of computer servers.
5, the cluster computer with distributed system for balancing loads as claimed in claim 1, wherein above-mentioned processor active task are an indivisible computing.
6, the cluster computer with distributed system for balancing loads as claimed in claim 1, computer server one is a load balance server wherein more than wherein above-mentioned, and these a plurality of front-end servers are wherein one in order to appoint processor active task to this load balance server, and this load balance server distributes in order to described node is carried out processor active task, be assigned to described node with described subtask, and described node is carried out load balance control this processor active task.
7, the cluster computer with distributed system for balancing loads as claimed in claim 1, wherein one of above-mentioned front-end server has more this cluster computer firewall functionality.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB021320276A CN1294509C (en) | 2002-09-06 | 2002-09-06 | Cluster computers possessing distributed system for balancing loads |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB021320276A CN1294509C (en) | 2002-09-06 | 2002-09-06 | Cluster computers possessing distributed system for balancing loads |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1480863A CN1480863A (en) | 2004-03-10 |
CN1294509C true CN1294509C (en) | 2007-01-10 |
Family
ID=34145084
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB021320276A Expired - Fee Related CN1294509C (en) | 2002-09-06 | 2002-09-06 | Cluster computers possessing distributed system for balancing loads |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN1294509C (en) |
Families Citing this family (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006004008A (en) * | 2004-06-15 | 2006-01-05 | Sony Computer Entertainment Inc | Processing control unit, computer system, distributed processing method, and computer program |
US7451347B2 (en) * | 2004-10-08 | 2008-11-11 | Microsoft Corporation | Failover scopes for nodes of a computer cluster |
EP1816565B1 (en) * | 2004-11-26 | 2011-06-08 | Fujitsu Ltd. | Computer system and information processing method |
US7464303B2 (en) * | 2005-06-09 | 2008-12-09 | International Business Machines Corporation | Autonomically adjusting configuration parameters for a server when a different server fails |
CN100394726C (en) * | 2005-06-30 | 2008-06-11 | 西安交通大学 | Method for raising reliability of software system based on strucural member |
JP2007304687A (en) * | 2006-05-09 | 2007-11-22 | Hitachi Ltd | Cluster constitution and its control means |
US7519855B2 (en) * | 2006-06-15 | 2009-04-14 | Motorola, Inc. | Method and system for distributing data processing units in a communication network |
CN101236513B (en) * | 2007-01-30 | 2012-07-18 | 阿里巴巴集团控股有限公司 | Distributed task system and distributed task management process |
CN101296176B (en) * | 2007-04-25 | 2010-12-22 | 阿里巴巴集团控股有限公司 | Data processing method and apparatus based on cluster |
CN101471810B (en) * | 2007-12-28 | 2011-09-14 | 华为技术有限公司 | Method, device and system for implementing task in cluster circumstance |
US8725875B2 (en) * | 2011-06-21 | 2014-05-13 | Intel Corporation | Native cloud computing via network segmentation |
CN102263824B (en) * | 2011-07-26 | 2017-07-18 | 深圳市中兴物联科技有限公司 | A kind of safe browsing method and virtual browser |
CN103365740B (en) * | 2012-04-06 | 2015-12-16 | 腾讯科技(深圳)有限公司 | A kind of data cold standby method and device |
CN102724311A (en) * | 2012-06-18 | 2012-10-10 | 苏州超集信息科技有限公司 | Workstation cluster system |
CN104461747B (en) * | 2014-12-31 | 2019-03-08 | 合一网络技术(北京)有限公司 | A kind of distributed task dispatching system |
CN110119984A (en) * | 2018-02-07 | 2019-08-13 | 青岛农业大学 | A kind of processing system for international trade tick financing |
CN109144783B (en) * | 2018-08-22 | 2020-08-18 | 南京壹进制信息科技有限公司 | Distributed massive unstructured data backup method and system |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1049307A1 (en) * | 1999-04-29 | 2000-11-02 | International Business Machines Corporation | Method and system for dispatching client sessions within a cluster of servers connected to the World Wide Web |
-
2002
- 2002-09-06 CN CNB021320276A patent/CN1294509C/en not_active Expired - Fee Related
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1049307A1 (en) * | 1999-04-29 | 2000-11-02 | International Business Machines Corporation | Method and system for dispatching client sessions within a cluster of servers connected to the World Wide Web |
Also Published As
Publication number | Publication date |
---|---|
CN1480863A (en) | 2004-03-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN1294509C (en) | Cluster computers possessing distributed system for balancing loads | |
US7328363B2 (en) | Dynamically configurable fault tolerance in autonomic computing with multiple service points | |
US6985952B2 (en) | Energy-induced process migration | |
US7444459B2 (en) | Methods and systems for load balancing of virtual machines in clustered processors using storage related load information | |
US20110119344A1 (en) | Apparatus And Method For Using Distributed Servers As Mainframe Class Computers | |
US20190294513A1 (en) | Adaptive multipath fabric for balanced performance and high availability | |
US9323628B2 (en) | Instance level server application monitoring, load balancing, and resource allocation | |
US8176501B2 (en) | Enabling efficient input/output (I/O) virtualization | |
US20070198982A1 (en) | Dynamic resource allocation for disparate application performance requirements | |
WO2013142991A1 (en) | Managing capacity on demand in a server cloud | |
WO2014122555A1 (en) | Workload balancing in distributed database | |
CN102096652A (en) | Method and apparatus to couple a module to a management controller on an interconnect | |
CN102158513A (en) | Service cluster and energy-saving method and device thereof | |
US20200042608A1 (en) | Distributed file system load balancing based on available node capacity | |
CN105242761A (en) | Server framework | |
US20170206110A1 (en) | Computer System for BMC resource management | |
US20140173332A1 (en) | Cascading Failover Of Blade Servers In A Data Center | |
CN107977271A (en) | A kind of data center's total management system load-balancing method | |
US20240080308A1 (en) | Automatic encryption for cloud-native workloads | |
CN103677993A (en) | Virtual machine resource load balancing system and method | |
CN116016028B (en) | Method, system, medium and equipment for creating multiple network interfaces for Pod based on IPvlan | |
Geyer et al. | Working with Disaggregated Systems. What are the Challenges and Opportunities of RDMA and CXL? | |
CN105046393A (en) | Cloud computing-based traffic resource management system | |
TW565801B (en) | Cluster computer having distributed load balancing system | |
CN108200109A (en) | Big data all-in-one machine |
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 | ||
C17 | Cessation of patent right | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20070110 Termination date: 20120906 |