CN101001246A - Method and system for estimating future network job cost - Google Patents
Method and system for estimating future network job cost Download PDFInfo
- Publication number
- CN101001246A CN101001246A CN 200610003669 CN200610003669A CN101001246A CN 101001246 A CN101001246 A CN 101001246A CN 200610003669 CN200610003669 CN 200610003669 CN 200610003669 A CN200610003669 A CN 200610003669A CN 101001246 A CN101001246 A CN 101001246A
- Authority
- CN
- China
- Prior art keywords
- grid
- provider
- tolerance
- cost
- grid provider
- 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.)
- Pending
Links
Images
Abstract
This invention provides a method for estimating cost of future grid operation by the reduction result of sorting grid operation and storing process operation, so that a grid customer proxy of the client system capable of submitting grid operation to the grid supplier computes the rate of the measurement based on the applied measurement and the measure of the grid supplier for processing specific grid operations, then the proxy sets up a list with an entry to compare the cost of the measurement based on the applied measurement and that measured by the supplier, then the proxy stores the list with the entry then tests the next grid operation, and the client proxy estimates the cost of the next operation processed by the supplier based on the specific number operated by the applied measurement.
Description
Technical field
The present invention relates generally to improved grid computing (grid computing), and be particularly related to effective client estimation of future network job cost.More especially, the present invention relates to compare by other network job cost history of having stored with this user of the current grid work of specific classification and this specific classification, the user who carries out future network job cost estimates.
Background technology
Connect since the first time between two computer systems, via being connected the just constantly development of new method that transmits data, resource and out of Memory between two computer systems.In typical network architecture, when two computer systems when connecting swap data, one of them of computer system is considered to the client of the request that sends and another is considered to handle the server end of request and return results.For improving the speed of the request of processing as possible, the size of server system and speed continue expansion.In addition, be to handle as possible the peak period when arrived a plurality of request each second, server system often is attached to together as one group, and asks to be dispensed between these servers in groups.Developed multiple method to server grouping, as troop, multisystem shared data (system combined body (sysplex)) environment, and business system.Utilize cluster of servers, a server typically is assigned management and distributes input request and output response.Other server typically parallel work-flow to handle request from the assignment of client.Like this, one of them of a plurality of servers in trooping can be served the request of client, and to be cluster of servers handling request and client can not detect.
Typically, server or server group run on the particular network platform as Unix or some Unix variant, and provide hosted environment to be used for the operation application.The function that each network platform can provide that scope is integrated from database, the service of trooping and fail safe are determined to work load management and problem.Each network platform typically provides different realizations, semantic behavior and API (API).
Yet only server being grouped into is a kind of limited method of improving the efficient of response time in the network with expanding treatment ability together.Like this, more and more, in company's network, not only server to be divided into groups, but server and server set of systems are organized as distributed resource.Have one effort that continues with cooperation between the external server of company's network internal and company's network, shared data, share the cycle and improve other interactive mode.In addition, also have one effort that continues with from the network of a company to the unessential element of service provider network outsourcing.And, also have a kind of trend, promptly in the control that is not subjected to same management system, coordinate resource-sharing but still can solve between the resource of problems such as fail safe, strategy, payment and membership.For example, the resource on personal desktop's computer typically is not subjected to the control of identical management system as the resource of company server cluster.Even the difference management group in company's network also may realize distinct management system.
To be from the problem that the available resources of the server that is positioned at zones of different, moves on the network platform in difference and other computing systems (having different security protocols and each security protocol by different management system control) are disperseed, cause developing the grid of the open standard that is used to operate grid environment.That grid environment is supported in is dynamic, distributed, share in the Virtual Organization and coordinate to use different resources.When the selection of resource (coming free different tissues to operate, have the system that disperses geographically of Different Strategies and management system) is organized into the processing operation request, in grid environment, create Virtual Organization.For example, grid manufacturer can develop grid environment, and buyer can submit to grid work to grid environment.
Grid manufacturer may be ready with different performances with handle grid work with different price strategy.Even standard (as those standards by the open standard tissue suggestion of grid) has defined standard monitoring, metering, estimates, accounts and has remembered the expense interface, grid manufacturer will still have different can be used for to handle the physical resource of grid work, and thus between different grid manufacturer price and performance still with different.In an example, grid manufacturer has to measure the employed grid of grid work manufacturer resource, and this will be referred to consider to comprise all multifactor complicated formulas of real resource use.For example, grid manufacturer may be exclusively used in specific processor resource specific operation, thereby except that the required actual processor amount of cycles of grid work, also at the special use of this processor this grid work is chargeed.
When grid manufacturer energy concentrate on the computing cycle rank to the actual use of physical resource monitor, when metering, cost sheet, grid client and user pay close attention to is that application on the application type rank is handled.As a result, getting in touch between the mode of observation of shortage grid user and grid manufacturer and grid work cost related.In addition, current each grid manufacturer still uses the monitoring of different physical resource measurement unit, metering and note expense grid work.Thus, because the grid manufacturer that client's grid work of application level and physical resource use is not related between measuring, the grid client be difficult to the different grid of comparison manufacturer the processing grid work cost and and estimate grid work is submitted to the future cost of same grid manufacturer thus.
Therefore, in view of foregoing, a kind of advantageous method, the system and program are provided, be used for grid work being divided into various classifications by linear module (metric unit) based on using with client definition, by the classification of grid work with the measure costs that is used to carry out grid work of grid manufacturer definition be converted to client definition based on the linear module cost of using, and the client definition changed of storage become the future cost of the grid work of original prediction identical category based on the linear module of using, estimate following operating cost.Especially, if earlier grid work epitome (microcosm) or less representative grid work are submitted to a plurality of grid manufacturer on less basis, to obtain the actual cost of each classification of grid work, the measure costs of grid manufacturer definition is converted to the linear module cost based on application of client definition, and client definition based on the tolerance rank comparative costs of using, large-scale grid work being submitted to the cost-efficient grid of tool manufacturer, is very favorable again.
Summary of the invention
In view of foregoing, the present invention provides improved grid computing substantially and provides the client future network job cost to estimate especially.More especially, the present invention relates to compare by other network job cost history of having stored with this user of the current grid work of specific classification and this specific classification, the user who carries out future network job cost estimates.
In one embodiment, make it possible to submit to grid work to calculate the ratio that is used to handle specific grid work based on tolerance of using and grid provider tolerance to the grid customer's representative of the client of the grid provider of being convenient to as required (on-demand) network environment.Then, the grid customer's representative creates a form with clauses and subclauses, compares the cost of measuring based on every grid provider of the tolerance of using and this grid provider based on the ratio that is calculated.Then, the grid customer's representative stores this form with these clauses and subclauses.Then, response detects next grid work, the grid customer's representative is based on the required specific quantity based on the metric operations of using of this next grid work, be converted into grid provider tolerance and multiply by every grid provider measure costs by this ratio, estimate that this grid provider handles the cost of this next grid work.
For calculating this ratio, at first, the grid customer's representative is distributed to grid provider with operation epitome (it is the less expression of specifiable lattice operation) and is used for handling at grid environment as required.In response to the result who receives this operation epitome and based on the expense of this operation epitome of processing of the grid provider tolerance of grid provider, then, the grid customer's representative calculates based on the ratio of the tolerance of using with grid provider tolerance, and the cost of measuring from the every supplier of cost identification who is used to handle this operation epitome.In addition, the grid customer's representative can at first have the job request that is used for this grid epitome requirement to the grid provider distribution, and is used to handle the cost of this grid epitome and the bid of other agreements (bid) from grid provider reception appointment.In addition, the grid customer's representative also can discern the cost of every grid provider tolerance in addition from the ratio that grid provider is announced.
When the grid customer's representative detects every grid provider measure costs through adjusting, no matter the grid customer's representative is by the price notice or by the charge to other grid works, upgrade the clauses and subclauses of this grid provider with every grid provider measure costs through adjusting, and reappraise the cost that this grid provider is handled next grid work automatically based on this every grid provider measure costs through adjusting, and do not need to recomputate ratio.
Described form can comprise the additional entries from the additional grid provider of the grid epitome of handling operations specific, and wherein said grid customer's representative is that each clauses and subclauses in the form are calculated the ratio based on the tolerance of the using tolerance different with each grid provider.When but a plurality of grid provider clauses and subclauses are the time spent in form, the grid customer's representative can estimate that each grid provider handles the cost of next grid work based on the every grid provider measure costs in this ratio and each clauses and subclauses, and relatively at required each cost that calculates based on the tolerance quantity operation of using of next grid work.
In addition, the clauses and subclauses in the form are according to the category classification of grid work.Thus, the grid customer's representative will detect next operation, grid work is sorted in one of grid work classification, and those clauses and subclauses of also classifying according to the identical category of grid work in the pro forma interview sheet.
Description of drawings
Being considered to novel feature of the present invention lists in appended claims.Yet, but when reading in conjunction with the accompanying drawings, by understanding the present invention self best, preferably use pattern and other target and advantage thereof with reference to the detailed description of following illustrative examples, wherein:
Fig. 1 has described an embodiment of the computer system that can realize and the present invention can realize therein in grid environment;
Fig. 2 is the block diagram of an embodiment of general type component in the signal grid environment scope;
Fig. 3 is the block diagram of an example of the signal architecture that can realize in grid environment;
Fig. 4 describes the flow chart that the result of the job request that the request of client submit job returns to grid provider and by grid provider bids;
Fig. 5 operation result's that to be signal client submit job return to grid provider and by grid provider flow chart;
Fig. 6 is a flow chart of describing the relative cost that the epitome be used to submit to grid work measures with the application of calculating every client definition to a plurality of grid providers;
Fig. 7 describes to be used for by the grid customer's representative's of the current grid work estimating future network job cost of specific classification block diagram relatively, and this specific classification has the cost historical record for other grid works storages of this specific classification.
Fig. 8 be illustrated to be used for grid work classification based on the cost of the application tolerance of client definition and form based on the cost example relatively of grid provider tolerance;
Fig. 9 describes to be used for the high-level logic flowchart of grid customer's representative according to the processes and procedures of the future cost of the grid work of the classification of the inventive method, system and program estimation client definition.
Embodiment
With reference now to accompanying drawing,, Fig. 1 has especially described an embodiment of the computer system that can realize and the present invention can realize therein in grid environment.Just as will be described further, grid environment comprises by management so that the multicomputer system of resource to be provided.In addition, as what also will further describe, the present invention may operate at various computing systems, comprises the multiple computing system, mobile system and the electronic installation that run on many different operating systems of managing in grid environment.
In one embodiment, computer system 100 comprises that bus 122 or other are used for the device of the information that transmits in computer system 100, and at least one is as the processing unit of processor 112, is connected to bus 122 and is used for process information.Bus 122 may comprise by bridge and being connected with adapter and by low latency and the Geng Gao stand-by period path of multibus controller in computer system 100 inner control.When realizing as server system, computer system 100 typically comprises a plurality of processors that are designed to improve network service capabilities.
In normal running, processor 112 is the general processors as the PowerPCTM processor of IBM, this processor deal with data and from such as the application software that obtains the dynamic storage device of random-access memory (ram) 114 and the static memory such as read-only memory (ROM) 116 under operating system control.Operating system can provide graphical user interface (GUI) for the user.In one embodiment, application software comprises the machine-executable instruction when the operation of describing in the flow chart of execution realization Fig. 9 and Figure 10 on processor 112 and other operation described here.Alternatively, step of the present invention may be carried out by the specific hardware components that comprises the hard-wired logic that is used to carry out this step, is perhaps carried out by any combination of programmed calculations thermomechanical components and custom hardware components.
The present invention can be used as computer program and provides, and comprises being stored in being used for computer system 100 with the executable instruction of the machine of carrying out processing procedure according to the present invention on the machine readable media.Term used herein " machine readable media " comprises that participation provides any medium of instruction to the assembly of processor 112 or other computer system that is used to carry out 100.This medium can be many kinds of forms, includes but not limited to, non-volatile media, easily loses medium and transmission medium.For example, the general type of non-volatile media comprise floppy disk, disk, hard disk, tape or any other magnetic medium, CD ROM (CD-ROM) or any other optical medium, punched card or any other have poroid pattern physical medium, programming ROM (PROM), can wipe any other medium that PROM (EPROM), electronics EPROM (EEPROM), flash memory, any other memory chip or cassette tape or computer system 100 could therefrom read and be suitable for store instruction.In the present embodiment, the example of a non-volatile media is the mass storage device 118 as the intraware that is described as computer system 100, but is understandable that also and can be provided by external device (ED).Easily the mistake medium comprises the dynamic memory as RAM114.Transmission medium comprises coaxial cable, copper cash or optical fiber, comprises the line of forming bus 122.Transmission medium also can be taked the form of sound wave or light wave, produces in radio frequency or infrared data transmission as those.
In addition, the present invention can be used as computer program and downloads, wherein, program command can be via being delivered to requesting computer system 100 with the data-signal of the network linking that is connected 134 of the communication interface 132 that is connected to bus 122 (for example, modulator-demodulator or network connect) by carrier wave or other propagation medium form from the remote dummy resource such as virtual resource 160.Virtual resource 160 can comprise from the virtual representation of the accessible resource of individual system or a plurality of systems, and wherein, perhaps each of multisystem all can be considered to run on the discrete set of the resource of independent platform, but to be coordinated by grid managers be virtual resource.Communication interface 132 provides the bidirectional data communication that is connected with network linking 134, and this network linking may be connected to, for example Local Area Network, wide area network (WAN) or be provided to the ISP (ISP) of the access of network 102.Especially, network linking 134 can be provided to the wired and/or wireless communication of one or more networks, and for example network 102, by using the virtual resource as virtual resource 160, is accessible when providing in grid environment 150.Grid environment 150 is parts of multi-type network, comprises point to point network, perhaps may be the part of single computer systems, as computer system 100.
As an example, network 102 relates to the use specific protocol, and as transmission control protocol (TCP) and Internet protocol (IP), the network that communicates to each other and the whole world of gateway are compiled.Network 102 uses electronics, electromagnetism or the optical signalling that carries digital data stream.On signal by diverse network and the network linking 134 and the signal by can carrying the communication interface 132 of numerical data back and forth 100 of computer systems be the exemplary form of the carrier wave of transmission information.Be understandable that, can realize the combination of optional network type, network and the foundation structure of network.
When realizing as server system, computer system 100 typically comprises via many Peripheral Component Interconnects (PCI) bus bridge to the accessible many communication interfaces of i/o controller.In this manner, computer system 100 allows to be connected to many network computers.
In addition, although do not describe, many external components and inner/outer device can join in the computer system 100, be connected to multi-controller, adapter with one of them expansion slot that is connected of multistage bus 122.For example, can be used as external components and add display unit, audio devices, keyboard or cursor control device.
It should be understood by one skilled in the art that the hardware of describing among Fig. 1 may change.In addition, it should be understood by one skilled in the art that described example and do not mean that the restriction of the relevant architecture of the present invention of hint.
With reference now to Fig. 2,, Fig. 2 has illustrated embodiment of general type component in the grid environment scope.In the present example, the assembly of grid environment 150 comprises the client 200 that is connected with grid environment system 240, and wherein grid environment system 240 is connected with cluster of servers 222, server 224, work station and desktop computer 226, data-storage system 228 and network 230.For the purpose of illustrating, the network type of the assembly in not describing the network site and being connected grid environment 150.Yet, should be understood that the assembly in the grid environment 150 can be positioned at the top layer of the network infrastructure architecture of utilizing overlapped multi-type network realization.The scope of network infrastructure can be from a plurality of large-scale enterprise system to Point-to-Point system to single computer systems.In addition, should be understood that the assembly in the grid environment 150 only is the representative of grid environment inner assembly type.Grid environment can be included in simply in the single computer systems and maybe can comprise in a plurality of business systems.
It should be understood that grid environment can be provided by grid manufacturer or supplier, wherein for example can be based on carrying out the cost that resource is used in required time quantum of grid work or the employed real resource amount computing grid environment 150.In addition, be understood that, grid environment 150 can comprise the gridding resource of supporting by as the single grid manufacturer of particular business enterprise or a plurality of grid manufacturer, wherein each manufacturer's persistent surveillance and the management manufacturer the resource group, but grid management system 240 can monitor the unconscious change of crossing all resources, and no matter is which resource which manufacturer provides.In addition, be understood that, although the resource discovering mechanism that is used to find the available grids resource is not shown, client 200 or grid management system 240 still can be found within the gridding resource of being advertised by this locality or the grid environment 150 or outside overall available directories.
Such as the focus target of the grid environment of grid environment 150 is to tissue and payment from the resource of a plurality of discrete systems that are considered as virtual resource 160.The system of client 200, cluster of servers 222, server 224, work station and desktop computer 226, data-storage system 228, network 230 and establishment grid management system 240 can be that different types of and regional distribution has independently management system.But through grid management system 240 by grid foundation structure can exchange message, resource and service.In addition, cluster of servers 222, server 224, work station and desktop computer 226, data-storage system 228 and network 230 are geographical goes up across country and the continent distributes or local access mutually.
In this example, client 200 is connected with grid management system 240.Client 200 can be represented any computer system that transmit a request to grid management system 240.Especially, client 200 can send dummy activity request (perhaps request for quote (RFQ)) and operation to grid management system 240.In addition, though the client 200 of present embodiment is described as using request visit grid environment 150, client 200 also may operate within the grid environment 150 in optional embodiment.
Although describe the system in the virtual resource 160 concurrently, in fact this system can be the part of system layer, and wherein some systems in the virtual resource 160 are local to client 200, and other system need be linked into external network.In addition, what emphasis was noted is that the system of describing in virtual resource 160 can physically be included in the client 200.
In order to realize grid environment 150, grid management system 240 has promoted mesh services.Mesh services can design according to the many body system structure, including, but not limited to, open grid service system (OGSA).Especially, grid management system 240 relates to by the link computing system to being characterised in that the management environment of creating a grid by the heterogeneous network environment of mesh services shared resource.
Especially, as will be with reference to figure 4-6 described, grid management system 240 can comprise and making it possible to from the automatic supervision of dynamic response bidding request and grid work and the mesh services of metering.In addition, grid management system 240 also can comprise the Additional Services that are used for the automatic function of execution in grid environment 150.
According to advantage of the present invention, client 200 comprises the grid customer's representative of the future cost that is used to estimate grid work.As will be with reference to figure 7 described, the grid customer's representative is the future cost of estimation grid work of all categories by the classification grid work, in form, store the cost of first consumption charge, and visit this form with future cost according to the operation of classification predicted grid for each classification of grid work.
With reference now to Fig. 3,, Fig. 3 has illustrated an example of the architecture that can realize in grid environment.As shown in the figure, architecture 300 comprises multi-functional layer.As will be described further, the present invention is can be in the process as realizing in one or more layers of the architecture of architecture 300, and this architecture can realize in the grid environment described in Fig. 2.That answers that emphasis notes is an example of the architecture that architecture 300 just can realize in the grid environment that the present invention can realize therein.What in addition, answer that emphasis notes is to realize the many body system structure in grid environment.
In each layer of architecture 300, at first, physics and logical resources layer 330 are organized the resource of the system in the grid.Physical resource is including, but not limited to server, storage medium and network.Logical resource is virtual and assemble physical layer to the available resources as operating system, disposal ability, internal memory, I/O processing, file system, data base administration person, memory manager, catalogue and other resource.
Next, web service layer 320 provides the interface between mesh services 310 and physics and the logical resource 330.Web service layer 320 realizes web service description language (sdl) (WSDL), Simple Object Access Protocol (SOAP) and the extend markup language (XML) of service interface including, but not limited to carrying out on Internet protocol (IP) or other transport network layer.As will be described further, when transport network layer returns an error code, web service layer 320 has strengthened having the XML message transaction of the SOAP envelope of having described error code, transmits to have the XML message transaction of error message to diagnostic agent.
At first, the use of resource management service 302 managing physicals and logical resource.Resource can including, but not limited to: handle resource, memory source and storage resources.The management of these resources comprises schedule job, distributes operation and is job management result's retrieval.The activity peak that resource management service 302 monitoring resources load and distribute operation not expect with balance resource load and absorption to the so busy part of grid.Especially, the user can specify preferred performance class so that resource management service 302 distributes operation to keep this preferred performance class in grid.
The second, the information between the computing system of information service 304 management in grid transmits and communicates by letter.Since can realize multi-communication protocol, information service 304 utilizes the communication of a plurality of networks of polytype management spanning of communication protocol.
The 3rd, the data in data management service 306 managing griddings transmit and storage.Especially, data management service 306 can mobile data node in the grid, will carry out the operation of request msg at this node.Can realize host-host protocol as the particular type of grid file host-host protocol (GridFTP).
At last, the articulamentum of the system in each runs on grid, fail safe service 308 is the Secure Application security protocol.The security protocol that fail safe service 308 can realize as open Secure Sockets Layer(SSL) is to provide safe transmission.In addition, fail safe service 308 can provide single signature mechanism, in case so that the user is by verifying, when just create and use letter of attorment when moving for the user carries out in grid.
Many services can be worked together so that several key functions of computing system to be provided.In first example, in grid, distributed calculation task.Data management service 306 is divided into different packet grid service requests with calculation task, and next this packet is distributed and management by resource management service 302.The result is collected by data management system 306 and consolidates.In second example, the storage resources of crossing over a plurality of computing systems in the grid is regarded as by data management service 306 management and the single virtual data-storage system that monitored by resource management service 302.
Application layer 340 comprises the application of using one or more mesh services available in the grid services layer 310.Advantageously, use via grid services layer 310 and web service 320 and be connected with logical resource 330, so that a plurality of alien system can interact and mutual operation with physics.
With reference now to Fig. 4,, Fig. 4 has described the flow chart of result's bid of the job request that client submit job request returns to grid provider and by grid provider.As shown in the figure, client 200 submit job requests 402 are to grid provider.Especially, each grid provider can be realized grid management system, the result's stream that is used for the bid processing procedure of management operations request and is used for management operations by the range of choice by the gridding resource of the grid provider of grid management system management.
In this example, the grid management system of grid provider comprises grid provider bidding request clauses and subclauses 404, receives job request and queuing at this.Grid provider bidding request clauses and subclauses guide each job request to workload calculator 408, the workload that its computational tasks request 402 is required.Especially, workload requires to comprise that for example operation is with the estimation of the computing cycle of needs and required hardware and software Platform Type.Workload calculator 408 amounts of sharing out the work are calculated as workload data 412 and are arrived cost calculation device 414 to bid formalization device and workload data 410.Cost calculation device 414 uses the current of workload calculating, job request requirement and resource use and is estimated the cost that original estimation is used for this grid work of processing operation request 402 appointments.Cost calculation device 414 round-trip cost data 416 are to bid formalization device 418.Bid formalization device 418 intensive work amount data 412 and cost data 416 are in bid 420, and bid 420 turns back to client 200 from grid provider.It is required accurately that bid 420 may be that grid work is carried out in promise, perhaps may comprise from the specification requirement in the job request 402 unusual, except with other changes.In addition, bid 420 can be considered the agreement of seeervice level, if specified the performance standard of the phone network lattice supplier promise of submitting to grid work to give grid provider subsequently.In addition, the price contract convention of reaching between grid client and the supplier before or after bid formalization device 418 can be arranged based on bid is set up bid.
With reference now to Fig. 5,, Fig. 5 has described the operation result's that the client submit job returns to grid provider and by grid provider flow chart.As shown in the figure, client 200 submit jobs 502 are to grid provider.Especially, as described in reference to Figure 4, each grid provider can be realized a grid management system, be used for by selecting gridding resource to come the managing gridding job stream to satisfy performance requirement for operation, the words that satisfy performance requirement if desired then monitor the operation process and adjust the operation resource, and the return results of managing client.
In this example, this grid management system of grid provider comprises job queue 504, and it receives operation 502 and can dispatch and assign operation 502 to blocked job 502 before the gridding resource at job scheduler 506.Especially, the bid that grid scheduler 502 is sent this operations specific based on this grid provider or to the transaction capabilities of the operation that receives from particular customer system access services level agreement (SLA) 508 that require to reach an agreement on, it comprises the performance requirement of operation 502.The gridding resource that grid scheduler 506 access process operations 502 are required, for example server A 516, server B 518 and server N 520.Although do not describe, the also addressable foundation of grid scheduler is used for other assemblies of the grid management system of grid work resource requirement, visits the resource from other grid environments, and if necessary gives other grid providers with the grid work distress sale.
In this example, grid scheduler 506 apportioned efforts 502 are for being assigned to 510,512 and 514 parts of server A 516, server B 518 and server N 520 respectively.Operation results management device 528 is concentrated respectively the result 522,524 and 526 from server A 516, server B 518 and server N520.Operation results management device 528 returns complete results 530 and gives client 200.In addition, operation results management device 528 upgrades accounting management device 532 when operation is finished.Accounting management device 532 is communicated by letter with workload manager (not describing), and workload manager monitors that the use of 502 pairs of server A 516 of operation, server B 518 and server N 520 is with the total workload of computational tasks 502 and the resulting cost of operation 502.Especially, if the factor that SLA 508 can specified control operation 502 total costs is the flotation costs yardstick and other price adjustment factors that are not met such as maximum cost, fixed cost performance requirement.
With reference now to Fig. 6,, Fig. 6 has described the epitome that is used to submit to grid work to the flow chart of a plurality of grid providers with the relative cost of the application tolerance of calculating every client definition.In this example, client 200 pro-rata grid works are epitome, and epitome is little, the representative operation than the macrolattice operation that need be submitted to grid provider.In this example, client 200 submits to grid work epitome 602 and grid work epitome 612 to grid provider 604 and 614 respectively.In one embodiment, client 200 has submitted to the job request of grid work epitome to grid provider 604 and 614, yet, as description with reference to figure 4, in optional embodiment, client 200 can for example price based on the verbal agreement band of ratio of announcing or contrast ratio and performance expectation submits to the grid work epitome to grid provider 604 and 614.
Each grid provider 604 and 614 is handled the identical mode of grid work with the grid provider described in figure 5 and is handled grid work epitome 602 and 612, and return results 606 and 616.As will further describing with reference to figure 7,8 and 9, client 200 is fetched the result and the cost of each grid work epitome 602 and 604, and the ratio of measuring based on the application tolerance and the grid provider of client definition is the calculating conversion value of each grid provider.Then, client 200 calculates the estimated cost of complete grid work according to the quantity of the operation of the application tolerance of the required client definition of the complete grid work of being adjusted by conversion value, the estimated cost of each grid provider relatively, and select the cost-efficient grid provider of tool.
According to an advantage, actual performance by each grid provider of sampling and cost and conversion cost are the application tolerance benchmark of client definition, client 200 can be before sending bigger grid work or a plurality of bigger grid work, based on the application of the client definition tolerance benchmark actual cost of performance relatively, rather than to the promise cost of performance.In this example, after the result and cost of each grid work epitome 602 and 612 that sampled, client 200 selects to send complete grid work 620 to grid provider 604, and grid work epitome 602 and 604 is representative set of grid work 602.Grid provider 604 is as handling grid work 620 and return results 624 to client 200 with reference to the description of figure 5.
With reference now to Fig. 7,, Fig. 7 has described and has been used for by the grid customer's representative's of the current grid work estimating future network job cost of specific classification block diagram relatively, and this specific classification has the cost historical record for other grid works storages of this specific classification.In this example, grid customer's representative 700 has described and has run on client 200 and maybe can realize software module within the other system of client 200.
As the description with reference to figure 6,702 controls of operation epitome controller are used to submit to the processing of the operation epitome of specifiable lattice operation to a plurality of grid providers.
Especially, can at first line up a plurality of grid providers of job request of operation epitome controller 702 with operation epitome.Except queuing has as the grid provider with reference to figure 6 described job request, can be with oral, in writing form, obtain other pricing information via mail or by some other electron exchange methods.Operation epitome controller 702 can be stored bid or ratio quotation, indicates the price of every computing cycle.In an example, first supplier returns price hourly, and second supplier returns the price of every floating-point operation, and the 3rd supplier returns the price based on the complicated formulas of supplier's grid environment.
Then, in case operation epitome controller 702 is obtained bid and ratio quotation from a plurality of grid providers, operation epitome controller 702 submit job epitomes are given a plurality of grid providers of selecting, and the operation epitome is that the less operation of submitting to than the macrolattice operation is represented.In an example, if needing every night, company on average merges 20,000,000 record, this operation epitome of distributing to selected grid provider so may comprise 1% of these records.In another example, the client does not send the part of actual grid operation, but submits the operation epitome of the simulation job that has the test program data to.
When having received the operation epitome controller 702 of the result of calculation that receives the operation epitome, then received from the expense of each grid provider to each operation epitome.In case operation epitome controller 702 also detects the operation epitome and is submitted to the time that the grid provider return results is spent after the grid provider.In this example, first supplier has spent 5 minutes return results and and $2.20, and second supplier spent 1 minute return results and and $1.74, and the 3rd supplier spent 2 minutes return results and and $3.40.Especially, be understood that, operation epitome controller 702 can receive cost information from each grid provider by multiple communication media, comprises from grid provider sending to client 200, E-mail communication, embedded record keeping number of tokens word signature separately and returning to client 200 together with handling receipt.
In case receive all every grid provider costs, cost comparator 706 compares the actual cost that grid provider is used to carry out specifiable lattice job class.Especially, cost comparator 706 calculates the application measure costs of the client definition of each grid work epitome.Especially, each grid provider is submitted the measure costs of supplier's definition to, for example cost or whenever based on the cost of supplier's complicated formulas per hour.Yet the client is at application level size definition grid work.For example, the tolerance of client definition is the cost that every record merges.In case cost comparator 706 has calculated the ratio of the application tolerance of client definition and grid provider tolerance, the ratio measured of the cost comparator 706 application tolerance that just can utilize client definition and grid is measured quantity with the application of the client definition of complete operation action required and is converted to price as conversion value so.Cost comparator 706 determine the cost-efficient grid provider of tool and trigger this grid work remainder or actual grid work be submitted to the cost-efficient grid provider of tool.
In addition, cost comparator 706 is stored in cost in the cost form 710 for the application tolerance of cost form 710 computing clients definition and grid provider measure costs ratio and with the application metric form of client definition.Shown in Figure 8 as reference, especially, cost form 710 comprises the application tolerance row 804 that are used for client definition.Although do not describe, the application tolerance of each client definition is relevant with the classification of client definition.The classification of client definition is the grid work kind that is used to carry out the similar operations type in the application layer classification.Grid work grader 704 definables classification or classification can be input and be specifiable lattice job class.In this example of implementing in Fig. 8, the classification of client definition is to use " merging in batches " and " the meteorological operation " of application tolerance of using the client definition of " each model iteration " of application tolerance of the client definition of " each merge ".It is the university that only carries out two computing functions that the application tolerance of these client definitions can for example be applied in the client.At first, the system of university carries out the pay sheet function that needs database to merge in batches.Secondly, the system of university carries out the intensive that needs more floating-point operations and calculates research work.It should be understood that and also can add other action types that need the resource that computing capability and other provide by as required grid provider that other classification is typically carried out by the client with classification.
The 4th row 810 of cost form 710 comprise conversion value, the ratio of the application tolerance of its expression client definition and grid provider tolerance.In this example, value cited under conversion value 810 comprises " per hour merging for 3,000,000 time ", " every synthesis unit merges for 600 times " and " every MFP computing merges for 2000 times ".Especially, conversion value is by quantity 706 calculating of cost comparator and that represent the application tolerance of the client definition that every grid provider tolerance ratio is realized.As previously described, can calculate conversion value based on the grid epitome.Yet, in other embodiments, also can calculate and the renewal conversion value based on complete operation epitome.
At last, the 5th row 812 of cost form 710 comprise the price asked of each grid provider tolerance.In this example, cited value comprises " Mei Xiaoshi $40 ", " whenever synthetic Dan Wei $0.02 " and " every MFP $0.08 " under the price asked 812 of each grid provider tolerance.In optional embodiment, supplier also can provide the nearest price of Price Range and supplier in the past.
According to advantage of the present invention, when the client wants to estimate the cost of a new grid work, grid work grader 704 these grid works of classification, and the cost that the application based on client definition of future cost estimator 708 searching cost forms 710 such other grid works is measured.Then, based on the application tolerance requirement of the client definition of this new grid work, future cost estimator 708 is estimated the cost of the grid work that this is new according to grid provider.In an example, based on the value of giving an example in the cost form 710, need 3,000, the new grid work that merges in batches for 000 time will spend $40 (per hour 3,000,000 merging/Mei Xiaoshi $40) on " acme grid ", on " wiley grid ", will spend $100 (600 merging/every He Chengdanyuan $.02 of every synthesis unit), on " coyote grid ", will spend $120 (2000 merging/every MFP $.08 of every MFP computing).
In addition, according to this advantage, if the cost that the application based on client definition of future cost estimator 708 searching cost forms 710 such other grid works is measured, but do not have usable record or cost out of date, future cost estimator 708 initiating task epitome controllers 702 are with the most current cost of the epitome of the specific classification classification of definite this grid work so.
In addition, what answer that emphasis notes is, because along with the past of time or the price of each grid provider tolerance of response market situation can change, can in every grid provider tolerance bid row 812 of cost form 710, insert new cost, and the price that refreshes the operation of classification in this classification estimates, and need not to change conversion value.For example, if the price of " acme grid " is brought up to Mei Xiaoshi $55, the price of " wiley grid " drops to every He Chengdanyuan $.013, and the price of " coyote grid " drops to " every MFP $0.05 ", do not change the application tolerance and the grid provider tolerance ratio of client definition cited under conversion value 810 row, so future cost estimator 708 can still estimate that based on the price after upgrading future cost is to finish 5 in the future, 000, the operation that merges in batches for 000 time, for example " acme grid " De $91.67 (per hour 3,000,000 merging/Mei Xiaoshi $55), " wiley grid " De $108.33 (600 merging/every He Chengdanyuan $.013 of every synthesis unit), and " coyote grid " De $125 (2000 merging/every MFP $.05 of every MFP computing).
With reference now to Fig. 9,, Fig. 9 has described and has been used for the high-level logic flowchart of grid customer's representative according to the processes and procedures of the future cost of the grid work of the classification of the inventive method, system and program estimation client definition.As shown in the figure, processing procedure starts from piece 900 and proceeds to piece 902 thereafter.It is the determining of classification of client definition that piece 902 has been described according to the qualitative classification grid work.Then, piece 904 has illustrated to determine that whether the grid work of same classification is at cost form acceptance of the bid valency.If at cost form acceptance of the bid valency, then process forwards piece 906 to the grid work of same classification.Piece 906 has been described the cost based on the previous current grid work of storing for the grid work of this particular category of price expectation, process end then in the cost form.Especially, the cost form is a cost form 710, and it comprises conversion value.Thus, the cost of predicting current grid work need be determined the quantity of the application metric operations of required client definition, and the quantity of the application metric operations of the conversion value conversion client definition by such other grid work be the cost of measuring based on grid provider.
Turn back to piece 904, if there is not the price of the grid work of this same classification in the cost form, processing procedure forwards piece 908 to so.In addition, although do not describe, even the price of the grid work of this same classification has been arranged, also can make the out-of-date decision of this price at piece 904 in the cost form, processing procedure forwards piece 908 to then.
Though, yet those skilled in the art will appreciate that and to it various changes and without departing from the spirit and scope of the present invention to be arranged in form and details with reference to the preferred embodiment special exhibition and described the present invention.
Claims (20)
1. computer implemented method that is used to estimate the future cost of grid work, described method comprises:
At the client place that makes it possible to the specifiable lattice operation is submitted to the grid provider of the grid environment that promotes to be used to handle described specifiable lattice operation, calculate based on the ratio of the tolerance of using with the grid provider tolerance that is used to handle described specifiable lattice operation;
Establishment has the form with the clauses and subclauses of the cost of the more described every grid provider tolerance based on tolerance of using and described grid provider of described ratio;
The described form that storage has the described clauses and subclauses that are used for described specifiable lattice operation; And
Response detects next grid work, based on being converted into described grid provider tolerance by described ratio and multiply by the specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required, estimate the cost of described next grid work of described grid provider processing.
2. according to the computer implemented method of future cost that is used to estimate grid work of claim 1, wherein also comprise making it possible to the client place that the specifiable lattice operation is submitted to the grid provider of the grid environment that promotes to be used to handle described specifiable lattice operation calculated based on the ratio of the tolerance of using with the grid provider tolerance that is used to handle described specifiable lattice operation:
The operation epitome of described specifiable lattice operation is distributed to described grid provider, in described grid environment, to handle; And
In response to the result who receives described operation epitome and based on expense from the described operation epitome of processing of the grid provider tolerance of described grid provider, calculate described based on the described ratio of the tolerance of using with described grid provider tolerance, and the cost of measuring from the described described every grid provider of cost identification that is used to handle described operation epitome.
3. according to the computer implemented method of future cost that is used to estimate grid work of claim 2, also comprise:
The job request that appointment is used to handle the requirement of described operation epitome is distributed to described grid provider; And
Response receives the bid that described grid provider is handled described operation epitome as requested, and described operation epitome is distributed to described grid provider to handle.
4. according to the computer implemented method of future cost that is used to estimate grid work of claim 1, wherein create the form that has with the clauses and subclauses of the more described cost of measuring based on every grid provider of tolerance of using and described grid provider of described ratio and also comprise:
Visit described every grid provider measure costs from least one of every grid provider tolerance rate that is used for handling the expense of described specifiable lattice operation and announcement that is received from described grid provider.
5. according to the computer implemented method of future cost that is used to estimate grid work of claim 1, also comprise:
The every grid provider measure costs through adjusting in response to detect described grid provider at described client place upgrades the described every grid provider measure costs in the described form, and
Based on being converted into described grid provider tolerance by described ratio and multiply by the described specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required, reappraise the cost that described grid provider is handled described next grid work automatically through adjusting.
6. according to the computer implemented method of future cost that is used to estimate grid work of claim 1,
Wherein at the client place that makes it possible to the specifiable lattice operation is submitted to the grid provider of the network environment that promotes to be used to handle described specifiable lattice operation, calculate based on the ratio of the tolerance of using and also comprise with the grid provider tolerance that is used to handle described specifiable lattice operation:
A plurality of operation epitomes of described specifiable lattice operation are distributed to a plurality of grid providers; And
In response to receiving a plurality of results and, calculates and describedly measure a plurality of ratios of each based on tolerance of using and the described a plurality of grid providers that are used for handling described specifiable lattice operation from a plurality of expenses of the described a plurality of operation epitomes of processing of each in described a plurality of grid providers;
Wherein creating the form have with the clauses and subclauses of the cost of the more described every grid provider tolerance based on tolerance of using and described grid provider of described ratio also comprises: the form that establishment has a plurality of clauses and subclauses, wherein each clauses and subclauses is the more described costs based on the tolerance of using and each described a plurality of grid provider tolerance of one of them described a plurality of grid provider, and described one of them described a plurality of grid provider have specific quota;
Wherein the described form that has described clauses and subclauses for described specifiable lattice job storage also comprises: each of the described a plurality of clauses and subclauses of storage in described form; And
Wherein estimate that based on being converted into grid provider tolerance by described ratio and multiply by the specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required the cost of described next grid work of described grid provider processing also comprises:
Based on described a plurality of costs of estimating described next operation of each processing of described a plurality of grid providers based on the tolerance of using; And
More described a plurality of cost is to discern the most cheap grid provider in the middle of described a plurality of grid provider.
7. according to the computer implemented method of future cost that is used to estimate grid work of claim 1, also comprise:
With the particular category of described specifiable lattice job class in a plurality of grid work classifications, wherein said particular category comprises that have can be by the grid work of described operation of measuring based on the tolerance of using;
Store the described clauses and subclauses of described specifiable lattice operation according to described particular category;
Response detects next grid work, according to described specifiable lattice job class described next grid work is classified;
In described form, search for the described clauses and subclauses that are associated with described identical particular category; And
According to the described clauses and subclauses of calculating based on described specifiable lattice operation, estimate that described grid provider processing and described specifiable lattice operation belong to the cost of described next grid work of identical particular category.
8. system that is used to estimate the grid work future cost, described system comprises:
The client that is connected with network service, wherein said client make it possible to the specifiable lattice operation is submitted to the grid provider that promotes to be used to handle the network environment of described specifiable lattice operation;
Described client also comprises:
Be used to calculate the device of the ratio of the grid provider tolerance of handling described specifiable lattice operation based on the tolerance and the described grid provider of application;
Be used to create the device that has with the form of the clauses and subclauses of the cost of the more described every grid provider tolerance based on tolerance of using and described grid provider of described ratio;
Be used to store the device of described form with the described clauses and subclauses that are used for described specifiable lattice operation; And
Be used to respond detection next grid work that grid is submitted to that is used for from described client, based on being converted into described grid provider tolerance by described ratio and multiply by the specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required, estimate the device of the cost of described next grid work of described grid provider processing.
9. the system that is used to estimate the grid work future cost according to Claim 8, the wherein said device that is used to calculate the ratio of the grid provider tolerance of handling described specifiable lattice operation based on the tolerance and the described grid provider of application, described client also comprises:
Be used for the operation epitome of described specifiable lattice operation is distributed to described grid provider, with the device of in described grid environment, handling; And
In response to the result who receives described operation epitome and based on expense from the described operation epitome of processing of the grid provider tolerance of described grid provider, be used to calculate described based on the described ratio of the tolerance of using with described grid provider tolerance, and the device of the cost of measuring from the described described every grid provider of cost identification that is used to handle described operation epitome.
10. according to the system that is used to estimate the grid work future cost of claim 9, wherein said client also comprises:
Be used for the job request that appointment is used to handle the requirement of described operation epitome is distributed to the device of described grid provider; And
Response receives the bid that described grid provider is handled described operation epitome as requested, is used for described operation epitome is distributed to the device of described grid provider to handle.
11. the system that is used to estimate the grid work future cost according to Claim 8, wherein saidly be used to create the device that has with the form of the clauses and subclauses of the cost of the more described every grid provider tolerance based on tolerance of using and described grid provider of described ratio, also comprise:
Be used for visiting the device of described every grid provider measure costs from least one of every grid provider tolerance rate that is used to handle the expense of described specifiable lattice operation and announcement that is received from described grid provider.
12. the system that is used to estimate the grid work future cost according to Claim 8, wherein said client also comprises:
In response to the every grid provider measure costs that detects described grid provider at described client place through adjusting, be used for upgrading the device of described every grid provider measure costs of described form, and
Be used for reappraising the device that described grid provider is handled the cost of described next grid work automatically based on being converted into described grid provider tolerance by described ratio and multiply by the described specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required through adjusting.
13. the system that is used to estimate the grid work future cost according to Claim 8, wherein said client also comprises:
Be used for a plurality of operation epitomes of described specifiable lattice operation are submitted to the device of a plurality of grid providers; And
In response to receiving a plurality of results and, be used for calculating described based on each the device of a plurality of ratios of tolerance of using and the described a plurality of grid providers tolerance that are used to handle described specifiable lattice operation from a plurality of expenses of the described a plurality of operation epitomes of processing of each in described a plurality of grid providers;
Be used to create described device with form of a plurality of clauses and subclauses, wherein each clauses and subclauses is the more described costs based on the tolerance of using and each described a plurality of grid provider tolerance of one of them described a plurality of grid provider, and described one of them described a plurality of grid provider have specific quota;
Be used for device in each of the described a plurality of clauses and subclauses of described form storage; And
Be used for handling the device of a plurality of costs of described next operation based on described each that estimate described a plurality of grid providers based on the tolerance of using; And
Be used for more described a plurality of cost to discern the device of grid provider the most cheap in the middle of described a plurality of grid provider.
14. the system that is used to estimate the grid work future cost according to Claim 8, wherein said client also comprises:
Be used for the device of described specifiable lattice job class to the particular category of a plurality of grid work classifications, wherein said particular category comprises that have can be by the grid work of described operation of measuring based on the tolerance of using;
Be used for storing the device of the described clauses and subclauses of described specifiable lattice operation according to described particular category;
Response detects next grid work, is used for according to the device of described specifiable lattice job class to described next grid work classification;
The device that is used for the described clauses and subclauses that are associated in the search of described form and described identical particular category; And
Be used for the described clauses and subclauses calculated according to based on described specifiable lattice operation, estimate that described specifiable lattice supplier process and described specifiable lattice operation belong to the device of cost of described next grid work of identical particular category.
15. a program that is used to estimate the future cost of grid work, described program package is contained in the computer-readable medium, and described program comprises computer executable instructions, and this instruction makes computer carry out the following step:
Calculating is based on the ratio of the grid provider tolerance of tolerance of using and the described specifiable lattice operation of specifiable lattice supplier process, and wherein said grid provider promotes to make it possible to handle the grid environment of described specifiable lattice operation;
Establishment has the form with the clauses and subclauses of the cost of the more described every grid provider tolerance based on tolerance of using and described grid provider of described ratio;
The described form that storage has the described clauses and subclauses that are used for described specifiable lattice operation; And
Response detects and is used for next grid work that grid is submitted to, based on being converted into grid provider tolerance by described ratio and multiply by the specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required, estimate the cost of described next grid work of described grid provider processing.
16. according to the program of future cost that is used to estimate grid work of claim 15, wherein said calculating also comprises the following steps: based on the step of the ratio that the tolerance of using and the grid provider of the described specifiable lattice operation of specifiable lattice supplier process are measured
The operation epitome of described specifiable lattice operation is distributed to described grid provider, in described grid environment, to handle; And
In response to the result who receives described operation epitome and based on expense from the described operation epitome of processing of the grid provider tolerance of described grid provider, calculate described based on the described ratio of the tolerance of using with described grid provider tolerance, and the cost of measuring from the described described every grid provider of cost identification that is used to handle described operation epitome.
17. the program of future cost that is used to estimate grid work according to claim 16 also comprises step:
The job request that appointment is used to handle the requirement of described operation epitome is distributed to described grid provider; And
Response receives the bid that described grid provider is handled described operation epitome as requested, and described operation epitome is distributed to described grid provider to handle.
18. according to the program of future cost that is used to estimate grid work of claim 15, the step that wherein said establishment has with the form of the clauses and subclauses of the cost of the more described every grid provider tolerance based on tolerance of using and described grid provider of described ratio also comprises the following steps:
Visit described every grid provider measure costs from least one of every grid provider tolerance rate that is used for handling the expense of described specifiable lattice operation and announcement that is received from described grid provider.
19. the program of future cost that is used to estimate grid work according to claim 15 also comprises step:
The every grid provider measure costs through adjusting in response to detecting described grid provider upgrades the described every grid provider measure costs in the described form, and
Based on being converted into described grid provider tolerance by described ratio and multiply by the described specific quantity based on the metric operations of using described every grid provider measure costs, that described next grid work is required, reappraise the cost that described grid provider is handled described next grid work automatically through adjusting.
20. the program of future cost that is used to estimate grid work according to claim 15 also comprises step:
With the particular category of described specifiable lattice job class in a plurality of grid work classifications, wherein said particular category comprises that have can be by the grid work of described operation of measuring based on the tolerance of using;
Store the described clauses and subclauses of described specifiable lattice operation according to described particular category;
Response detects next grid work, according to described specifiable lattice job class described next grid work is classified;
In described form, search for the described clauses and subclauses that are associated with described identical particular category; And
According to the described clauses and subclauses of calculating based on described specifiable lattice operation, estimate that described specifiable lattice supplier process and described specifiable lattice operation belong to the cost of described next grid work of identical particular category.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200610003669 CN101001246A (en) | 2006-01-11 | 2006-01-11 | Method and system for estimating future network job cost |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200610003669 CN101001246A (en) | 2006-01-11 | 2006-01-11 | Method and system for estimating future network job cost |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101001246A true CN101001246A (en) | 2007-07-18 |
Family
ID=38693055
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200610003669 Pending CN101001246A (en) | 2006-01-11 | 2006-01-11 | Method and system for estimating future network job cost |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101001246A (en) |
-
2006
- 2006-01-11 CN CN 200610003669 patent/CN101001246A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101080736B (en) | Computer execution method and system for automatically controlling grid job distribution | |
| US7571120B2 (en) | Computer implemented method for estimating future grid job costs by classifying grid jobs and storing results of processing grid job microcosms | |
| JP4954089B2 (en) | Method, system, and computer program for facilitating comprehensive grid environment management by monitoring and distributing grid activity | |
| US7562035B2 (en) | Automating responses by grid providers to bid requests indicating criteria for a grid job | |
| US7788375B2 (en) | Coordinating the monitoring, management, and prediction of unintended changes within a grid environment | |
| CN101013427B (en) | Method and system for managing data | |
| Zhao et al. | Flexible service selection with user-specific QoS support in service-oriented architecture | |
| Jarvis et al. | Performance prediction and its use in parallel and distributed computing systems | |
| US20060149652A1 (en) | Receiving bid requests and pricing bid responses for potential grid job submissions within a grid environment | |
| EP2405386A2 (en) | Brokered cloud computing architecture | |
| CN1771479A (en) | Method and system of configuring elements of a distributed computing system for optimized value | |
| Djemame et al. | Brokering of risk‐aware service level agreements in grids | |
| Li et al. | A survey of resource provisioning problem in cloud brokers | |
| US7925755B2 (en) | Peer to peer resource negotiation and coordination to satisfy a service level objective | |
| US9152937B2 (en) | Message sequence management of enterprise based correlated events | |
| Deldari et al. | A survey on preemptible IaaS cloud instances: challenges, issues, opportunities, and advantages | |
| Haji et al. | A SNAP-based community resource broker using a three-phase commit protocol: A performance study | |
| CN101001246A (en) | Method and system for estimating future network job cost | |
| US20060106689A1 (en) | Method and apparatus for documenting a contribution of a remotely accessed computing resource to a recipient organization | |
| Smith et al. | An economic model for medical data retrieval using mobile software agents | |
| Buyya et al. | The gridbus middleware for market-oriented computing | |
| Kantere | A framework for cost-aware cloud data management | |
| Kandru | Economy based resource allocation in grids | |
| Farrag | Scheduling Mechanism for Mobile Computing | |
| Elmeleegy | Faculty of Engineering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20070718 |