US20090138551A1 - Method of Managing Workloads and Associated Distributed Processing System - Google Patents

Abstract

Various embodiments enable a server system to consider a variety of priorities in determining how to organize, allocate, and manage client workloads in a distributed processing system. In at least one embodiment, a method may include generating a capability vector for one or more distributed devices, the capability vector defining a capability of the one or more distributed devices, determining a client priority for one or more client system, the client priority defining an order that a client system is granted access to submit one or more workloads to a distributed processing system for processing, assigning a workload priority to one or more workloads, the workload priority defining an order in which a workload is submitted to the distributed processing system for processing, and distributing one or more workloads form a client system to one or more distributed devices in response to the capability vector, the client priority, and the workload priority.

Description

REFERENCE TO EARLIER APPLICATIONS
• This is a continuation of and claims priority to U.S. patent application Ser. No. 09/794,969, entitled “System and Method for Monetizing Network Connected User Bases Utilizing Distributed Processing Systems,” filed on Feb. 27, 2001, which is incorporated by reference.
• This application is a continuation-in-part application of and claims priority to the following applications: U.S. Pat. No. 6,654,783 entitled “System and Method for Monetizing Network Connected user Bases Utilizing Distributed Processing Systems;” U.S. Pat. No. 6,891,802 entitled “Network Site Testing Method and Associated System;” U.S. patent application Ser. No. 09/539,023 entitled “Sweepstakes Incentive Model and Associated System;” U.S. patent application Ser. No. 09/539,448 entitled “Capability Based Distributed Parallel Processing System and Associated Method;” U.S. patent application Ser. No. 09/539,428 entitled “Method of Managing Distributed Workloads and Associated System;” U.S. patent application Ser. No. 09/539,107 entitled “Distributed Backup System and Associated Method;” and U.S. patent application Ser. No. 09/603,740 entitled “Method of Managing Workloads and Associated Distributed Processing System;” each of which was filed on Mar. 30, 2000 and each of which is incorporated by reference.
• This application is also a continuation-in-part of and claims priority to the following: U.S. Pat. No. 7,020,678 entitled “Machine Generated Sweepstakes Entry Model and Associated Distributed Processing System;” U.S. Pat. No. 6,963,897 entitled “Machine Generated Sweepstakes Entry Model and Associated Distributed Processing System;” and U.S. patent application Ser. No. 09/602,844 entitled “Data Conversion Services for Associated Distributed Processing System;” each of which was filed on Jun. 23, 2000 and each of which is incorporated by reference.
BACKGROUND
• A number of traditional Internet-focused businesses have based their ability to monetize users via an advertising model. Given the current decline in rates for web page banner advertising, as well as rates for other web page advertising forms, these businesses are often unable to develop and implement sustainable business models that lead to profitability. In many cases, these Internet-focused businesses provide a service in exchange for advertising exposure which their users are willing to allow because they value the service being offered. Unfortunately, most of these business face revenue rates (e.g., advertising rates) that are declining faster than their costs (e.g., cost of services provided to users).
SUMMARY
• This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the subject matter, nor is it intended to be used to limit the scope of the subject matter.
• Various embodiments enable a server system to consider various priorities in determining how to organize, allocate, and manage client workloads in a distributed processing system.
• In at least one embodiment, a method of operating a distributed processing system may include generating a capability vector for each distributed device, generating a client priority for each client system, and assigning a client workload priority to each client workload. Then, distributing a workload to the distributed devices of the distributed processing system in response to the capability vector, the client priority, and the workload priority.
• In a further embodiment, a computer readable storage media may include instructions to monitor a resource of a distributed device, generate a status for the distributed device, and modify the status when resources are added, deleted, and/or availability of a resource changes. Finally, the computer readable storage media may include instructions to send the status to a server system in response to a status request.
BRIEF DESCRIPTION OF THE DRAWINGS
• The same numbers are used throughout the drawings to reference like features.
• FIG. 1A is a block diagram for a distributed processing system having client capability and incentive features, according to one or more embodiments.
• FIG. 1B is a block diagram for information flow among customer systems, server systems and client systems, according to one or more embodiments.
• FIG. 2A is a block diagram for a client system, according to one or more embodiments.
• FIG. 2B is a block diagram for processing elements within a client system, according to one or more embodiments.
• FIG. 2C is a block diagram for a client system agent installed on a client system, according to one or more embodiments.
• FIG. 2D is an example user interface for a client system agent, including incentive advertising, according to one or more embodiments.
• FIG. 3A is a block diagram for server systems, according to one or more embodiments, including a control system, a sweepstakes system and a workload database.
• FIG. 3B is a block diagram for servers systems, customer systems, client systems and outsourced host systems, according to one or more embodiments.
• FIG. 3C is a block diagram for a server system processor, according to one or more embodiments.
• FIG. 3D is an alternative block diagram for a server system processor, according to one or more embodiments.
• FIG. 4 is a functional block diagram for an example sweepstakes incentive operation according to one or more embodiments.
• FIG. 5A is a block diagram for a distributed processing system for a network site indexing application, according to one or more embodiments.
• FIG. 5B is a functional block diagram for an indexing operation according to one or more embodiments.
• FIG. 6A is a block diagram for a server system according to one or more embodiments, including a control system, a workload database, and a database of client capabilities balancing vectors.
• FIG. 6B is a functional block diagram for client capabilities balancing of workloads according to one or more embodiments.
• FIG. 7A is a block diagram for a distributed processing system, according to one or more embodiments, including example network sites on which site testing is to be conducted, such as load testing and/or quality-of-service testing.
• FIG. 7B is a functional block diagram for site-testing, according to one or more embodiments.
• FIG. 8 is a block diagram of a distributed processing system for a data backup application, according to one or more embodiments.
• FIG. 9 is a block diagram of an alternative representation of an interconnection fabric for a distributed processing system environment, according to one or more embodiments.
• FIG. 10 is a block diagram of a more detailed block diagram for a client system agent installed on a client system, according to one or more embodiments.
• FIG. 11A is a more detailed flow diagram for machine generated sweepstakes entries according to one or more embodiments.
• FIG. 11B is an alternative detailed flow diagram for machine generated sweepstakes entries according to one or more embodiments.
• FIG. 12A is a block diagram of a distributed processing system that allows customers to select client system attributes, according to one or more embodiments.
• FIG. 12B is a block flow diagram for client system attribute selection, according to one or more embodiments.
• FIG. 13A is a block diagram of a distributed processing system that provides data conversion services, according to one or more embodiments.
• FIG. 13B is a block flow diagram for data conversion services within a distributed processing system, according to one or more embodiments.
• FIG. 14A is a block diagram of a distributed processing system that provides data transmission caching, according to one or more embodiments.
• FIG. 14B is a block diagram of a distributed processing system that provides data sharing and file distribution, according to one or more embodiments.
• FIG. 15 is a block diagram of an alternative representation for a distributed processing system, according to one or more embodiments.
• FIG. 16 is a block diagram of a representation for a distributed processing system including security subsystems, according to one or more embodiments.
• FIG. 17A is a block diagram of a client system and server systems communication interface, according to one or more embodiments.
• FIG. 17B is a block diagram of communication layers for client system and server systems communication, according to one or more embodiments.
• FIG. 18A is a detailed block diagram for an embodiment of security activities for server systems.
• FIG. 18B is a detailed block diagram for an embodiment of security activities for client systems.
• FIG. 19 is a block diagram for a distributed processing system and environment in which network service providers are enabled to monetize their user bases in accordance with one or more embodiments.
• FIG. 20 is a block diagram representing the components for a client agent along with a representative indication of responsibility for those components in accordance with one or more embodiments.
DETAILED DESCRIPTION
• The present embodiments contemplate the identification of the capabilities of distributed devices connected together through a wide variety of communication systems and networks and the aggregation of these capabilities to accomplish processing, storage, broadcasting or any other desired project objective. For example, distributed devices connected to each other through the Internet, an intranet network, a wireless network, home networks, or any other network may provide any of a number of useful capabilities to third parties once their respective capabilities are identified, organized, and managed for a desired task. These distributed devices may be connected personal computer systems (PCs), internet appliances, notebook computers, servers, storage devices, network attached storage (NAS) devices, wireless devices, hand-held devices, or any other computing device that has useful capabilities and is connected to a network in any manner. Various embodiments further contemplate providing an incentive, which may be based in part upon capabilities of the distributed devices, to encourage users and owners of the distributed devices to allow the capabilities of the distributed devices to be utilized in the distributed parallel processing system of the present embodiments.
• The number of usable distributed devices contemplated by various embodiments can be very large. Unlike a small local network environment, for example, as may be used by an Internet Service Provider (ISP), which may include less than 100 interconnected computers systems to perform the tasks required by the ISP, the various embodiments can utilize a multitude of widely distributed devices to provide a massively distributed processing system. With respect to the various embodiments, a multitude of distributed devices refers to greater than 1,000 different distributed devices. With respect to the various embodiments, widely distributed devices refer to a group of interconnected devices of which at least two are physically located at least 100 miles apart. With respect to the various embodiments, a massively distributed processing system is one that utilizes a multitude of widely distributed devices. The Internet is an example of a interconnected system that includes a multitude of widely distributed devices. An intranet system at a large corporation is an example of an interconnected system that includes multitude of distributed devices, and if multiple corporate sites are involved, may include a multitude of widely distributed devices. A distributed processing system according to the various embodiments that utilize such a multitude of widely distributed devices, as are available on the Internet or in a large corporate intranet, is a massively distributed processing system according to the various embodiments.
• FIG. 1A is a block diagram for a distributedparallel processing system100 according to at least some embodiments. Thenetwork102 is shown having a cloud outline to indicate the unlimited and widely varying nature of the network and of attached client types. For example, thenetwork102 may be the Internet, an internal company intranet, a local area network (LAN), a wide area network (WAN), a wireless network, a home network or any other system that connects together multiple systems and devices. In addition,network102 may include any of these types of connectivity systems by themselves or in combination, for example, computer systems on a company intranet connected to computer systems on the Internet.
• FIG. 1A also showsclient systems108,110 . . .112 connected to thenetwork102 throughcommunication links118,120 . . .122, respectively. In addition,server systems104,other systems106, andcustomer systems152 are connected to thenetwork102 throughcommunication links114,116 and119, respectively. The client system capabilities block124 is a subset of theserver systems104 and represents a determination of the capabilities of theclient systems108,110 . . .112. The incentives block126 is also a subset of theserver systems104 and represents an incentive provided to the users or owners of theclients systems108,110, . . .112 for allowing capabilities of theclients systems108,110 . . .112 to be utilized by the distributedprocessing system100. Theclient systems108,110 and112 represent any number of systems and/or devices that may be identified, organized and utilized by theserver systems104 to accomplish a desired task, for example, personal computer systems (PCs), internet appliances, notebook computers, servers, storage devices, network attached storage (NAS) devices, wireless devices, hand-held devices, or any other computing device that has useful capabilities and is connected to a network in any manner. Theserver systems104 represent any number of processing systems that provide the function of identifying, organizing and utilizing the client systems to achieve the desired tasks.
• The incentives provided by the incentives block126 may be any desired incentive. For example, the incentive may be a sweepstakes in which entries are given toclient systems108,110 . . .112 that are signed up to be utilized by the distributedprocessing system100. Other example incentives are reward systems, such as airline frequent-flyer miles, purchase credits and vouchers, payments of money, monetary prizes, property prizes, free trips, time-share rentals, cruises, connectivity services, free or reduced cost Internet access, domain name hosting, mail accounts, participation in significant research projects, achievement of personal goals, or any other desired incentive or reward.
• As indicated above, any number of other systems may also be connected to thenetwork102. Theelement106, therefore, represents any number of a variety of other systems that may be connected to thenetwork102. Theother systems106 may include ISPs, web servers, university computer systems, and any other distributed device connected to thenetwork102, for example, personal computer systems (PCs), internet appliances, notebook computers, servers, storage devices, network attached storage (NAS) devices, wireless devices, hand-held devices, or any other connected computing device that has useful capabilities and is connected to a network in any manner. Thecustomer systems152 represents customers that have projects for the distributed processing system, as further described with respect toFIG. 1B. Thecustomer systems152 connect to thenetwork102 through thecommunication link119.
• It is noted that the communication links114,116,118,119,120 and122 may allow for communication to occur, if desired, between any of the systems connected to thenetwork102. For example,client systems108,110 . . .112 may communicate directly with each other in peer-to-peer type communications. It is further noted that the communication links114,116,118,119,120 and122 may be any desired technique for connecting into any portion of thenetwork102, such as, Ethernet connections, wireless connections, ISDN connections, DSL connections, modem dial-up connections, cable modem connections, fiber optic connections, direct T1 or T3 connections, routers, portal computers, as well as any other network or communication connection. It is also noted that there are any number of possible configurations for the connections fornetwork102, according to one or more embodiments. Theclient system108 may be, for example, an individual personal computer located in someone's home and may be connected to the Internet through an Internet Service Provider (ISP).Client system108 may also be a personal computer located on an employee's desk at a company that is connected to an intranet through a network router and then connected to the Internet through a second router or portal computer.Client system108 may further be personal computers connected to a company's intranet, and theserver systems104 may also be connected to that same intranet. In short, a wide variety of network environments are contemplated by one or more embodiments on which a large number of potential client systems are connected.
• FIG. 1B is a block diagram forinformation flow150 amongcustomer systems152,server systems104, andclient system134, according to one or more embodiments. Theserver systems104, as discussed above, may include any number of different subsystems or components, as desired, including client system capabilities block124 and incentives block126. Theserver systems104 send project andbenchmark workloads130 toclient systems134. A benchmark workload refers to a standard workload that may be used to determine the relative capabilities of theclient systems134. A project workload refers to a workload for a given project that is desired to be completed. The project workload may be, for example, a workload for projects such as network site content indexing, network site testing including network site load testing and network site quality of service testing, data back-up, drug design, drug interaction research, chemical reaction studies, bioinformatics including genetic and biological analyses, human genome analyses, pair-wise comparisons including fingerprint and DNA analyses, data mining, Internet hosting services, intranet hosting services, auction services, market clearing services, payment systems, bio-informatic simulations, knowledge management services, trading services, data matching services, graphics rendering, or any other desired project.
• Client systems134, as discussed above, may be any number of different systems that are connected to theserver systems104 through anetwork102, such asclient systems108,110 . . .112 inFIG. 1A. Theclient systems134 sendresults132 back to theserver systems104 after theclient systems134 complete processing any given workload. Depending upon the workload project, theserver systems104 may then provide results156 tocustomer systems152. Thecustomer systems152 may be, for example, an entity that desires a given project to be undertaken, and if so, provides the project details anddata158 to theserver systems104.
• FIG. 2A is a block diagram for anexample client system108 according to at least one embodiment. In this simplified block diagram, anoriginal workload204 is received throughline208 from aninterface206. Theoriginal workload204 represents a portion of the processing, storage or other activity required to complete the desired task for which theserver system104 is trying to accomplish. Thisoriginal workload204 is sent by theserver system104 through thenetwork102 and received by theclient system108 throughcommunication link118. Theclient system108 processes theoriginal workload204. Followingline212, results202 are then stored for transferring alongline210 to interface206.Interface206 may then communicate the results back to theserver system104 throughcommunication line118, or to other client systems (for example, with peering of client systems) and then through thenetwork102.
• It is noted that the workload received byclient system108 and the processing or activity performed may depend up a variety of factors, as discussed further below. In part, this workload allocated by theserver system104 to eachclient system108,110 and112 may depend upon the capabilities of the client system, such as the processing power, disk storage capacity, communications types, and other capabilities available from the various components of the systems within theclient system108.
• Theserver systems104 can select the workloads for theclient system108 and may control when these workloads are performed, through operational code (i.e., an agent) residing and installed on theclient system108. Alternatively, the owner or user of theclient system108 may determine when workloads are procured or obtained from theserver systems104, as well as when these workloads are performed, for example, by accessing theserver systems104 through thenetwork102. For example, the seversystem104 may download to theclient system108 upon request one or more workloads. At the same time, an agent residing on theclient system108 may operate to process the workload or multiple workloads downloaded to theclient system108. It is noted, therefore, that the agent may be simultaneously managing more than one workload for any number of projects. When the workload is complete, the agent may inform the owner or user of theclient system108 the results are ready to be communicated back. Theclient system108 may then upload results to theserver system104 and download new workloads, if desired. Alternatively, these logistical and operational interactions may take place automatically through control of the agent and/or theserver systems104.
• FIG. 2B is a block diagram for processing elements within aclient system108 according to at least some embodiment. In this diagram,client system108 is contemplated as a personal computer. In a personal computer, aninternal bus260 would typically have a variety of different devices connected to it. For example, aCPU250 could be connected through thebus260 to a 20video processor252, a floating point processor254 (often integrated within the CPU itself), and digital signal processors (DSPs), such as those found on sound cards and modems. In addition, any of a variety ofother processing devices258 may be included. Furthermore, other types of devices may be connected, such ashard drives264, which provide disk storage capabilities, and adigital camera262.
• It is noted, therefore, that the capabilities forclient systems108,110 . . .112 may span the entire range of possible computing, processing, storage and other subsystems or devices that are connected to a system connected to thenetwork102. For example, these subsystems or devices may include: central processing units (CPUs), digital signal processors (DSPs), graphics processing engines (GPEs), hard drives (HDs), memory (MEM), audio subsystems (ASs), communications subsystems (CSs), removable media types (RMs), and other accessories with potentially useful unused capabilities (OAs). In short, for any given computer system connected to anetwork102, there exists a variety of capabilities that may be utilized by that system to accomplish its direct tasks. At any given time, however, only a fraction of these capabilities are typically used on theclient systems108,110 . . .112. The present embodiment can take advantage of these unused capabilities.
• It is also noted that along with receiving the workload, theclient system108 will also receive an agent that manages the completion of the workload. This agent may be software that is customized for the particular computer system and processing capabilities of theclient system108. For example, if the client system is a personal computer as shown inFIG. 2B, the agent may be a program that operates in the background of the computer's operating system. When the agent determines that there is unused processing or other capabilities, the agent may take advantage of it. For example, if the user is using a word processing application to create a document, little processing power is being utilized by the word processing program, leaving the computer's CPU and video processor underutilized. Thus, the agent could execute commands to these processors during dead cycles. In this way, the agent may facilitate the completion of workload processing in a reduced time. In addition, this agent may be self-updating upon connecting to theserver systems104, so that the agent may be kept up to date with current software revisions and workload activities. It is also noted that the agent may manage work on multiple workloads at the same time, so that any given distributed device connected to thenetwork102 may be working on a plurality of workloads at any given time.
• FIG. 2C is a block diagram for an exampleclient system agent270. Theagent270 may include asecurity subsystem272 that controls the interface of theclient system108 with theagent270. Thesecurity subsystem272 may help keep the workloads secure and may help to keep theclient systems108 from suffering any security problems in completing the workload. For example, theagent272 may operate to keep viruses from attacking theclient system108 while theclient system108 is processing the workload through the operation of the agent. Thesecurity subsystem272, therefore, may provide the interface for theworkloads130 and theresults132.
• The client'ssystem agent270 may also include aworkload engine274, a statistics/user interface/incentive advertising block276, and a workload package and updateprocessing block278. In the example shown inFIG. 2C,workloads130 pass through thesecurity subsystem272 and alongline280 to the workload package and updateprocessing block278. In thisblock278, theagent270 may be updated by theserver systems104. Alternatively, theagent270 may determine, when connected to theserver systems104, whether it should be updated and then accomplish that updating automatically. Once the workload package is processed, theworkload engine274 may receive theworkload following line288. Theworkload engine274 works on the workload, ultimately completing the workload. The results or status of the workload may then be sent through thesecurity subsystem272 followingline282. Theresults132 may then be provided back to theserver systems104.
• The statistics/user interface/incentive advertising block276 may provide workload, incentive and other statistics, as well as any other desired interface features, to the user of the client system. For example, theblock276 may show a user the expected amount of processing time it will take for the client system to complete a workload task based upon the capabilities of the system. As also shown, theblock276 may receiveinformation following lines286 and284 from the workload package and updateprocessing block278 and from theworkload engine274. If desired, security information from thesecurity subsystem272 could also be displayed to the user of the client system. It is noted that the information displayed to the user of the client system may be modified and selected as desired without departing from the claimed subject matter.
• With respect to incentive advertising, theblock276 may also show the user of the client system how this processing time might change depending upon various possible upgrades to the capabilities of the client system, such as a faster microprocessor, more memory, more disk storage space, etc. Furthermore, the client system capabilities may be shown correlated to the incentives provided to the client system for participation. Thus, the user may be provided information as to how the user's incentives would increase or change depending upon other computer systems or upgraded capabilities the user could acquire. This incentive value increase may also be tied to upgrades to particular vendor's devices. For example, if the user's device is a computer system having an ABC microprocessor, theblock276 may provide the user information as to increased incentive values based upon an upgrade to a more powerful ABC microprocessor. Similarly, if the user's device is a computer system obtained from ABC, theblock276 may provide the user information as to increased incentive values based upon an upgrade to a more powerful ABC computer system.
• FIG. 2D is anexample user interface276 for a client system agent, including incentive advertising, according to one or more embodiments. In the example shown,interface276 is awindow230 that may be displayed on a distributed device, for example, a computer system. Thiswindow230 displays the desired information for the agent client manager. As indicated above, this agent client manager is initially downloaded from theserver systems104 and thereafter may be updated at various times when the client system is communicating with the server systems. Theinterface276, as shown, includesinterface tabs221,222,224,226,228,244,246, and248. These interface tabs may be selected through the user of a pointing device or keyboard attached, for example, to a computer system graphically displaying thewindow230. It is noted that theinterface tabs221,222,224,226,228,244,246, and248 are only examples, and the number, arrangement and content of tabs may be modified as desired. In addition, theexample user interface276 depicted inFIG. 2D is only an example and may be modified as desired.
• InFIG. 2D, the processorvalues interface tab224 is the one currently selected by the user. Theprocessor values tab224 includes example information that may be displayed to the user. Assuming that a workload is being processed by the agent client manager, the user may select the button242 (Show My Incentive Values) to show the user's current incentive values associated with the workload being performed. The personal incentive values chart232 (My Personal Incentive Values) may then be displayed to the user. As shown, the incentive values are provided in a relative scale from 1 to 10. The key designation240 represents the incentives associated with the users current central processing unit (CPU) or microprocessor.
• As indicated above, this incentive information may also be tied to the specific vendor of the user's CPU, for example, ABC Company's CPU. Thus, as shown, the key designation240 (My current processor) and the correspondingbar graph portion236 represent incentives for the user's current CPU (e.g., a 166 MHz processor). Thekey designation238 represents the incentives that the user is projected to have if the user were to upgrade the CPU. In this manner, a user may be provided an incentive to increase the capabilities of the distributed device, and a vendor may be provided advertising so that the user is also directed to a particular upgrade.
• Looking further toFIG. 2D, other similar incentive related information tabs may be provided for any desired capability of the distributed device. For example, tab246 (Memory Values) represents information that may be provided for the memory capabilities of the distributed device. Tab224 (Graphics Values) represents information that may be provided for the graphics capabilities of the distributed device. Tab226 (Communications Values) represents information that may be provided for the communication capabilities of the distributed device. Tab228 (Storage Values) represents information that may be provided for the storage capabilities of the distributed device. Tab248 (System Values) represents information that may be provided for the system capabilities as a whole for the distributed device.
• In addition to these incentive related information tabs, other tabs may be included to provide information and control for any desired features of the agent client manager. For example, the tab244 (Current: Prime Search) represents information that may be displayed to the user about the current workload being performed by the agent client manager, for example, a search for large prime numbers. The tab221 (Settings) represents information that may be displayed to the user about various settings for the client agent manager. In particular, thetab221 may provide the user the ability to control any desired aspect of the operation of the agent client manager. For example, the user may be able to select a portion of the capabilities that may be utilized (e.g., a maximum of 20% of the system memory), the types of workloads that may be performed (e.g., only scientific research projects), the times when the agent may utilize system resources (e.g., only between 12 to 6 am, or only when the system is idle), or any other desired operational feature. It is noted that in addition to upgrade incentive information indicated above, the user may also be provided information as to how incentives would increase if the user allocated or changed the settings for the agent client manager.
• This user selection of operational features allows for workloads to be scheduled or balanced based upon user input and desires. These user vectors, as indicated above, would allow users to dedicate their device capabilities to specific research projects (e.g., cancer, Parkinson's disease, Internet, genetics, space science, etc.), to specific non-profit or for profit organizations (e.g., Greenpeace, Celera, etc.), educational institutions (e.g., University of Texas), a specific group of likeminded users, or any other entity or endeavor. This affiliation selection allows the distributed processing system to automatically include a user's device capabilities in a pool dedicated to the chosen affiliation. Additionally, a user could choose to mix various percentages and allocations of device capabilities among multiple affiliations. It is noted that the user does not have to make any affiliation selection and may not allocate 100 percent of device capabilities. Rather, only a portion of the device capabilities may be allocated to a particular affiliation, leaving the remainder non-allocated and not affiliated. The capability allocation may also be a system-wide (i.e., course) allocation, such as some desired percent of overall device capabilities. The capabilities allocation may also be subsystem specific (i.e., fine) allocation, such as allocation of particular subsystem capabilities to particular affiliations.
• Now looking toFIG. 3A, theserver systems104 may be one or more computer systems that operate to identify client system capabilities, organize workloads, and utilize client systems to accomplish a desired task. Theserver systems104 includes a control system304 aworkload database308, and asweepstakes system306, as discussed more below. Theworkload database308 stores any desired project task, which may be broken up into discrete workload tasks WL1, WL2 . . . WLN, as represented byelements336,338 . . .340. The workload database may also store one or more benchmark workloads (BWL)335 that may be utilized to determine client system capabilities in response to a standard workload. Throughline312, theworkload database308 communicates withcontrol system304.Control system304, for example, receivesoriginal workload322 and transfers it to theinterface320 throughline330. Theinterface320 then transfers theworkload322 to thenetwork102 throughline114. Thisworkload322 is ultimately received asworkload204 byclient system108,110 or112, as shown inFIG. 2A. Theresult324 is ultimately received by thecontrol system304 throughinterface320 andline328.
• In allocating workloads, thecontrol system304 may consider the capabilities of theclient systems108,110 and112 to which thecontrol system304 is sending workloads. For example, ifclient108 has more processing power thanclient110, thecontrol system304 may allocate and send more difficult or larger workloads. Thus,client108 may receiveWL1336 andWL2338, whileclient110 would only receive WL3. Alternatively, theworkload database308 could be organized with differing levels of processing power or capability requirements for each workload. In this way,WL1336 may represent a greater processing or system capability requirement thanWL2338. It should be noted that workload may be a processing task, a data storage task, or tied to any other of a variety of capabilities that may be utilized on theclient systems108,110 . . .112.
• As indicated above, to encourage owners or users of client systems to allow their system capabilities to be utilized bycontrol system104, an incentive system may be utilized. This incentive system may be designed as desired. Incentives may be provided to the user or owner of the clients systems when the client system is signed-up to participate in the distributed processing system, when the client system completes a workload for the distributed processing system, or any other time during the process. In addition, incentives may be based upon the capabilities of the client systems, based upon a benchmark workload that provides a standardized assessment of the capabilities of the client systems, or based upon any other desired criteria.
• One example use of a benchmark workload is to use the benchmark workload to determine incentive values. For example, theserver systems104 may be designed to send out a standard benchmark workload once an hour to eachclient system108,110 . . .112. If a client system is not available at that time for any reason, the workload would not be completed by the client system, and there would be no incentive value generated for that client system. In this example, the benchmark workload may be a timed work-set that would exercise each subsystem with capabilities within the client system that was desired to be measured. A more capable client system would then generate greater incentive values from executing the benchmark workload, as compared to a lesser capable client system. These incentive values may be utilized as desired to determine what the client system should get in return for its efforts. For example, if the incentive were a sweepstakes as discussed further below, the number of entries in the sweepstakes may be tied to the system's performance of the benchmark workload. Thus, the faster or better the client system performs the benchmark workload, the more entries the client system would receive.
• In the embodiment shown inFIG. 3A, theserver systems104 may include asweepstakes system306 that functions withcontrol system304 to provide incentives for the users or owners ofclient systems108,110 and112 to allow their system capabilities to be used by theserver systems104. Thecontrol system304 may determine asweepstakes entry value302 that is sent alongline310 to thesweepstakes system306. Thesweepstakes system306 may then receive sweepstakes entry332 and provide it to thesweepstakes engine330 throughline334. Thesweepstakes engine330 may process the entries and determine a winner, when desired. In the embodiment shown, therefore, entries to the sweepstakes may be generated each time a unit of work is accomplished by one or more of the subsystems within aclient system108,110 or112 via an agent installed on the device for the purposes of managing and completing units of work. The total entries for any period of time would, therefore, be dynamic depending on how many are received. Odds of winning would then be determined by the total number of entries received and the total number of entries contributable to any given entrant.
• FIG. 3B is another example block diagram of a distributedprocessing system300 includingservers systems104,customer systems152,client systems134 and out-sourcedhost systems340, according to one or more embodiments. Theservers systems104 may include ananalytic subsystem346, a results/workload production subsystem344, aproject pre-processing subsystem342, a client agent subsystem243, and an incentive advertising subsystem245. The incentive advertising subsystem245 may operate to provide advertising information, for example, the upgrade incentive information as discussed with respect toFIG. 2D. The client agent subsystem243 may operate to download an agent to theclient systems134 and to update this agent at times when theserver systems104 are communicating with theclient systems134.
• Thecustomer systems152, which represent customers that have projects that they desired to be processed by the distributed processing system, may be connected to theproject preprocessing subsystem342 to provide projects to theservers systems104. These projects are processed by theproject pre-processing subsystem342 and passed to the results/workloads production subsystem344, which produces and sends outworkloads130 and receives back results130. Theanalytic system346 then takes the results and processes them as desired. Completed project information may then be provided from theanalytic system346 to thecustomer systems152. In this manner, the projects of thecustomer systems152 may be processed and project results reported by the distributed processing system in one or more embodiments. Also, as shown, theworkloads130 and theresults132, or other tasks of theserver systems104, may be processed and handled by out-sourcedhost systems340, if desired. Thus, some or all of theworkloads130 may be sent first to out-sourcedhost systems340. Out-sourcedhost systems340 then sendworkloads130A to theclient systems134 and receiveback results132A. The out-sourcedhost systems340 then send theresults132 back to theserver systems104. It is noted that this out-sourcing of server system tasks may be implemented as desired for any given task that theserver systems104 may have. It is further noted that, if desired, theserver systems104 may perform all of the desired functions of theserver systems104 so that no out-sourcedhost systems340 would be used.
• FIG. 3C is a block diagram for one embodiment of aserver system processor350, according to one or more embodiments. Anagent abstraction layer360 may sendworkloads130 and receiveresults132. Thesecurity subsystem354 may interact with theagent abstraction layer360 and provide information to adata parser352 and an application programming interface (APIs)block356. TheAPIs block356, thedata parser352 and aworkload manager558 may interact to accomplish the desired tasks for theserver system processor350. It is noted that for this embodiment, the API protocol could be controlled and provided to other host systems.
• FIG. 3D is an alternative block diagram for aserver system processor350, according to one or more embodiments. In this embodiment, the APIs block356 and theagent abstraction layer360 are not present. Thedata parser352, theworkload manager358 and thesecurity subsystem354 interact to provide the desired server system tasks. It is noted that for this embodiment, the security subsystem is controlled and utilized for communicating with client systems.
• FIG. 4 is a functional block diagram for asweepstakes operation400 by thesystem server104 according to an embodiment. Inblock402, theserver systems104 may sign-up client systems in “accept clients”block402. Followingline418, theserver systems104 identify the capabilities of the client's computer and processing systems in the “determine client system capabilities”block404. Control passes alongline420 to the “distribute workloads to client systems”block406, where theserver systems104 allocate workloads to eachclient system108,110 and112. This workload may also be a benchmark workload, as indicated above, that acts as an entry workload to determine the entries or entry values for the client system. As also indicated above, in distributing the workloads inblock406, theserver system104 may take into consideration the capabilities of the client systems to which workloads are being distributed. Theclient systems108,110 and112 then operate to complete the workloads allocated to them. Theserver system104 receives back workload results in “receive workload results”block408.
• At this point, control passes alongline424 to the “determine sweepstakes entries”block410. In thisblock410, theserver system104 determines the entry value for the workload completed or for a standard benchmark or entry workload completed. This entry value may be weighted upon a variety of factors including factors such as the amount of work completed, the difficulty level of the processing required, and the accuracy of the results. It is noted that any desired weighting may be utilized. Thus, it is understood that a wide variety of considerations may be utilized to determine the entry value weighting for the sweepstakes.
• Although the weighting determination is shown inblock410 inFIG. 4, the entry value may also be determined, in whole or in part, when a client system signs on to the distributed processing distributed system of one or more embodiments. For example, if a client system has state-of-the-art CPU, video processor, DSP engine, memory, and large amounts of free disk storage space, a high entry value may be allocated to this client system up-front. In contrast, a client system that has a slow CPU, a weak video processor, no DSP engine, little memory, and little free disk storage space may be allocated a small entry value. In this way, the owners or users of the client systems may be provided immediate feedback as to the potential sweepstakes entry value of their computer systems, devices and system capabilities.
• It is further noted that the entry value may take any desired form and may be, for example, a multiplier that will be used for each unit of workload completed. In this way, the owner or user will readily be cognizant that a state-of-the-art system will yield a high multiplier, where as an older system, system capability or device will yield a low multiplier. Such feedback, whether communicated to the owner or user immediately upon signing up or upon completion of each workload, will create an incentive for owners and/or users to acquire state-of-the-art systems, thereby further increasing the potential processing power of the distributed processing system of one or more embodiments. In addition, different workload projects may be designated with different entry values, as well. For example, some workload projects may require particular hardware or software processing systems within a client system or device. Thus, the number of client systems that are capable of performing the task would be limited. To further encourage participation by those owners or users with capable systems, the entry value for taking on particular workloads and/or systems with the desired features may be allocated higher entry values.
• Referring back toFIG. 4, control passes alongline426 to the “process entries”block412. In thisblock412, the sweepstakes entries are processed and stored as desired. Followingline428, “end of entry period”decision block414 represents a determination of whether the time for getting entries into the sweepstakes has ended. If not, the control continues to line430 and back toblocks402,404 and/or406, depending upon what is desired. Once the entry period has ended, control flows alongline432 to “determine winners”block416. Theserver system104 then identifies from among the entries, who the winning client system or systems will be.
• The entry period may be any desired time frame and may include multiple overlapping time frames, as desired. For example, winners may be determined daily for entries each day, monthly for entries within a month, and/or yearly for entries within one year. In addition, special entry periods may be generated, if desired, for example where a particularly important workload project had a short time frame in which it is to be completed.
• FIGS. 1A-B,2A-C,3A-D, and4 are directed to example embodiments for a distributed processing system according to one or more embodiments, including a sweepstakes reward or incentive feature, as shown in the embodiments ofFIG. 3A andFIG. 4.
• FIGS. 6A and 6B further describe a capabilities scheduling feature, in which theserver systems104 may identify and consider any of a variety of client system capability vectors in determining how to organize, allocate and manage workloads and projects.FIGS. 5A and 5B describe a distributed processing system and workload project that accomplishes network site indexing.FIGS. 7A and 7B describe a distributed processing system and a workload project that accomplishes network site testing, such as quality of service (QoS) testing and load testing. AndFIG. 8 describes a distributed processing system with respect to a corporate intranet, that accomplishes distributed data back-up.
• FIG. 9 is an alternative representation for the interconnection fabric for a distributed processing system environment and describes idle client system identification and shared component client systems.FIG. 10 describes a client system agent installed on a client system.FIGS. 11A and 11B further describe machine generated sweepstakes entries.FIGS. 12A and 12B describe client capability selection features.FIGS. 13A and 13B describe data conversion services.FIG. 14A describes a distributed processing system that provides data transmission 20 caching.FIG. 14B describes a distributed processing system that provides data sharing and file distribution functions. AndFIG. 15 describes an alternative representation for a distributed processing system, according to one or more embodiments.
• Looking now toFIG. 5A, block diagram is depicted of a distributedprocessing system550 for a network site indexing application, according to one or more embodiments. As stated above with respect toFIG. 1A, thenetwork102 may be a wide variety of networks. For this network site indexing application, thenetwork102 may be the Internet having a multitude ofnetwork sites552 . . .554. Eachnetwork site552 . . .554 may have a variety of different content types that may be indexed, ranging from complex sites to relatively simple sites. For example,network site552 includestext570A,images570B,audio streams570C,video streams570D, files570E and other content570F.Network site554 is less complex and includestext572A,images572B, andother content572C. Bothnetwork sites552 and554 are connected to thenetwork102 throughcommunication lines558 and556, respectively.
• As discussed above, theserver systems104 manage workloads for theclient systems108,110 . . .112. Theclient systems108,110 . . .112 process these workloads and produce indexing results. The resulting index may be stored at a centrally managed site, such as centralindex storage block560, or may itself be distributed over the possibly millions of indexingclients108,110, . . .112, as shown by remote index storage blocks562,564, . . .566. If remote index storage is utilized, a master database content index may be stored locally, for example, in the centralindex storage block560. This content index may then direct relevant searches to the distributed massively parallel engine for search queries.
• Referring now toFIG. 5B, a functional block diagram is shown for a networksite indexing operation500 according to one or more embodiments. As described inFIG. 1A with respect toother systems106, there may be any number of computer and processing systems connected to thenetwork102. Any one of theseothers systems106 may publish information on thenetwork102 for access by any other system connected to thenetwork102. This information to be indexed may take a wide variety of forms, including, for example, text, images, audio streams, video streams, databases, spreadsheets, PDF files, Shockwave data, Flash data, applications, data files, chat streams, or any other information, data or data streams that may be accessible on a network site. The distributed processing system of one or more embodiments may have as a workload the task of indexing this potentially massive amount of information.
• For example, where thenetwork102 is the Internet or a large intranet, a large amount of processing power and time can be used to create an accurate, complete and up-to-date index of the information. The Internet uses an IP (Internet Protocol) address protocol to direct traffic around the Internet. The IP address is the address of a computer attached to a TCP/IP (Transmission Control Protocol/Internet Protocol) network. Every system on the network must have a unique IP address. IP addresses are typically written as four sets of numbers separated by periods. The TCP/IP packet uses 32 bits to contain the IP address, which is made up of a network and host address (NETID and HOSTID). The more bits used for network address, the fewer remain for hosts. Web pages within a particular web site with a unique address may be addressed through URLs (Uniform Resource Locator) associated with that web site. In short, there is a limited, but very large, number of possible IP addresses for uniquely identifiable Internet sites that may be accessed and analyzed to generate an index of Internet sites and web pages via URLs.
• The operation diagram ofFIG. 5B starts with the “clients receive indexing workloads”block502. In this block, thesystem server104 provides theclients systems108,110 . . .112 with a workload task to index a portion of the information accessible on thenetwork102. For example, with the Internet, each workload may be single EP address or groups of URLs or, in some cases, large data types contained on single sites or pages. Followingline514, the “clients interact with other systems”block504 represents the operation of the agent installed on theclient systems108,110, . . .112 to access the network sites, according to the assigned workload, and index the information accessible on that site. This indexing may include all types of information accessible on that site, including text, audio, image, video, etc.
• Next, followinglines516 and518, theclient systems108,110, and112 completes the workload tasks, get the results ready for transmission, and sends those results back to thesystem server104 in “clients complete workload”block506 and “indexing results sent to server system”block508. Control passes alongline520 to “index compiled for use”block510 where the server system formats and/or compiles the results for use. For example, the index results may be utilized for accurate, complete and up-to-date search information for thenetwork102. As indicated with respect toFIG. 5A, the resulting index may be stored remotely or locally followingline522. Thus,element524 represents remote storage of the index andelement526 represents central storage of the index. It is noted that the index may also be stored with a mixture of central and remote storage, as desired. In addition, as indicated above, a directory or summary index for the resulting index may be generated and stored centrally, if desired. It is further noted that the summary index may be stored in any other desired fashion, for example, it may be distributed and stored on a number of client systems.
• FIG. 6A is a block diagram for aserver system104 according to an embodiment, including acontrol system304, aworkload database308, and a database ofcapability vectors620. Theworkload database308 includes a variety of sets of workload projects WL1, WL2 WLN. For each workload project, there may be multiple workload units. For example, workload project WL1 includes workload units WL11, WL12 . . . WL1N, as represented byelements640,642 . . .644, respectively. Similarly, workload project WL2 includes workload units WL21, WL22 . . . WL2N, as represented byelements646,648, . . .650, respectively workload project WL3 includes workload units WL31, WL32 . . . WL3N, as represented byelements652,654 . . .656, respectively.
• It may be expected that different workload projects WL1, WL2 . . . WLN within theworkload database308 may require widely varying processing requirements. Thus, in order to better direct resources to workload projects, the server system may access various system vectors when a client system signs up to provide processing time and other system or device capabilities to the server system. This capability scheduling helps facilitate project operation and completion. In this respect, thecapability vector database620 keeps track of any desired feature of client systems or devices incapability vectors CBV1, CBV2 . . . CBVN, represented byelements628,630 . . .632, respectively. These capability vectors may then be utilized by thecontrol system304 throughline626 to capability balance workloads.
• This capability scheduling according to one or more embodiments, therefore, allows for the efficient management of the distributed processing system. This capability scheduling and distribution will help maximize throughput, deliver timely responses for sensitive workloads, calculate redundancy factors when appropriate, and in general, help optimize the distributed processing computing system of one or more embodiments. The following TABLE 1 provides lists of capability vectors or factors that may be utilized. It is noted that this list is an example list, and any number of vectors or factors may be identified and utilized, as desired.

• TABLE 1
  Example Client Capability Vectors orFactors

  1

  BIOS Support:

  a.

  BIOS Type (brand)

  	b.	ACPI	c.	S1, S2, S3, and S4 sleep/wake
  				states
  	d.	Dl, D2 and D3 ACPI device states	e.	Remote Wake Up Via Modem
  	f.	Remote Wake Up Via Network	g.	CPU Clock control
  	h.	Thermal Management control	i.	Docked/Undocked state control
  	j.	APM 1.2 support	k.	Hotkey support
  	1.	Resume on Alarm, Modem Ring	m.	Password Protected Resume from
  		and LAN		Suspend
  	n.	Full-On power mode	o.	APM/Hardware Doze mode
  	p.	Stand-by mode	q.	Suspend to DRAM mode
  	r.	Video Logic Power Down	s.	HDD, FDD and FDC Power Down
  	t.	Sound Chip Power Down	u.	Super I/O Chip Power Down

  2

  CPU Support:

  a.

  CPU Type (brand)

  	b.	MMX instruction set	c.	SIMD instruction set
  	d.	WNI instruction set	e.	3DNow instruction set
  	f.	Other processor dependent	g.	Raw integer performance
  		instruction set(s)
  	h.	Raw FPU performance	i.	CPU LI data cache size
  	j.	CPU Ll instruction cache size	k.	CPUL2 cache size
  	1.	CPU speed (MHz/GHz . . . )	m.	System bus (MHz/GHz . . . ) speed
  				supported
  	n.	Processor Serial Number	o.	CPUID

  3

  Graphic Support

  a.

  Graphics type (brand)

  	b.	# of graphics engines	c.	Memory capacity
  	d.	OpenGL support	e.	Direct3D/DirectX support
  	f.	Color depth supported	g.	MPEG VII decode assist
  	h.	MPEG1/II encode assist	i.	OS support
  	j.	Rendering type(s) supported	k.	Single-Pass Multitexturing support
  	1.	True Color Rendering	m.	Triangle Setup Engine
  	n.	Texture Cache	o.	Bilinear/Trilinear Filtering
  	p.	Anti-aliasing support	q.	Texture Compositing
  	r.	Texture Decompression	s.	Perspectively Correct Texture
  				Mapping
  	t.	Mip-Mapping	u.	Z-buffering and Double-buffering
  				support
  	v.	Bump mapping	w.	Fog effects
  	x.	Texture lighting	y.	Video texture support
  	z.	Reflection support	aa.	Shadows support

  4

  Storage Support

  a.

  Storage Type (brand)

  	b.	Storage Type (fixed, removable,	c.	Total storage capacity
  		etc.)
  	d.	Free space	e.	Throughput speed
  	f.	Seek time	g.	User dedicated space for current
  				workload
  	h.	SMART capable

  5

  System

  a.

  System Type (brand)

  	b.	System form factor (desktop,
  		portable, workstation, server, etc.)

  6

  Communications Support

  a.

  Type of Connection (brand of ISP)

  	b.	Type of Connection Device (brand	c.	Hardware device capabilities
  		of hardware)
  	d.	Speed of connection	e.	Latency of connection
  	f.	Round trip packet time of	g.	Number of hops on connection type
  		connection
  	h.	Automatic connection support	i.	Dial-up only (yes/no)
  		(yes/no)
  	j.	Broadband type (brand)	k.	Broadband connection type
  				(DSL/Sat./Cable/T 1/Intranet/etc.)

  7	Memory	a.	Type of memory error correction
  			(none, ECC, etc.)

  	b.	Type of memory supported (EDO,	c.	Amount of total memory
  		SDRAM, RDRAM, etc.)
  	d.	Amount of free memory	e.	Current virtual memory size
  	f.	Total availablevirtual memory size

  8

  Operating System

  a.

  Type of operating system (brand)

  b.

  Version of operating system

  c.

  Health ofoperating system

  9.	System application software	a.	Type of software loaded and/or
  			operating on system

  	b.	Version of software	c.	Software features enabled/disabled
  	d.	Health of software operation

• FIG. 6B is a functional block diagram for capabilities determination andscheduling operation600 for workloads in a distributed processing system according to an embodiment. Initially, various vectors are identified for which capability information is desired in the “identify client system capability vectors”block602. Followingline612, theserver systems104 then capability balances workloads amongclient systems108,110 and112 based upon the capability vectors in the “capability scheduling workloads based on vectors”block604. Then the capabilities scheduled workloads are sent to theclient systems104 for processing in the “send capability scheduled workloads”block606.
• This capability scheduling and management based upon system related vectors allows for efficient use of resources. For example, utilizing the operating system or software vectors, workloads may be scheduled or managed so that desired hardware and software configurations are utilized. This scheduling based upon software vectors may be helpful because different software versions often have different capabilities. For example, various additional features and services are included in MICROSOFT WINDOWS '98® as compared with MICROSOFT WINDOWS '95®. Any one of these additional functions or services may be desired for a particular workload that is to be hosted on a particular client system device. Software and operating system vectors also allow for customers to select a wide variety of software configurations on which the customers may desire a particular workload to be run. These varied software configurations may be helpful, for example, where software testing is desired. Thus, the distributed processing system of one or more embodiments may be utilized to test new software, data files, Java programs or other software on a wide variety of hardware platforms, software platforms and software versions. For example, a Java program may be tested on a wide proliferation of JREs (Java Runtime Engines) associated with a wide variety of operating systems and machine types, such as personal computers, handheld devices, etc.
• From the customer system perspective, the capability management and the capability database, as well as information concerning users of the distributed devices, provide a vehicle through which a customer may select particular hardware, software, user or other configurations, in which the customer is interested. In other words, utilizing the massively parallel distributed processing system of one or more embodiments, a wide variety of selectable distributed device attributes, including information concerning users of the distributed devices, may be provided to a customer with respect to any project, advertising, or other information or activity a customer may have to be processed or distributed.
• For example, a customer may desire to advertise certain goods or services to distributed devices that have certain attributes, such as particular device capabilities or particular characteristics for users of those distributed devices. Based upon selected attributes, a set of distributed devices may be identified for receipt of advertising messages. These messages maybe displayed to a user of the distributed device through a browser, the client agent, or any other software that is executing either directly or remotely on the distributed device. Thus, a customer may target particular machine specific device or user attributes for particular advertising messages. For example, users with particular demographic information may be targeted for particular advertisements. As another example, the client agent running on client systems that are personal computers may determine systems that are suffering from numerous page faults (i.e., through tracking operating system health features such as the number of page faults). High numbers of page faults are an indication of low memory. Thus, memory manufactures could target such systems for memory upgrade banners or advertisements.
• Still further, if a customer desires to run a workload on specific device types, specific hardware platforms, specific operating systems, etc., the customer may then select these features and thereby select a subset of the distributed client systems on which to send a project workload. Such a project would be, for example, if a customer wanted to run a first set of simulations on personal computers with AMD ATHLON® microprocessors and a second set of simulations on personal computers with INTEL PENTIUM III® microprocessors. Alternatively, if a customer is not interested in particular configurations for the project, the customer may simply request any random number of distributed devices to process its project workloads.
• Customer pricing levels for distributed processing may then be tied, if desired, to the level of specificity desired by a particular customer. For example, a customer may contract for a block of 10,000 random distributed devices for a base amount. The customer may later decide for an additional or different price to utilize one or more capability vectors in selecting a number of devices for processing its project. Further, a customer may request that a number of distributed devices be dedicated solely to processing its project workloads. In short, once device attributes, including device capabilities and user information, are identified, according to one or more embodiments, any number of customer offerings may be made based upon the device attributes for the connected distributed devices. It is noted that to facilitate use of the device capabilities and user information, capability vectors and user information may be stored and organized in a database, as discussed above.
• Referring now toFIG. 12A, a block diagram depicts a distributedprocessing system1200 that allows customers to select client system attributes, such as device capabilities and user characteristics, according to one or more embodiments. In this embodiment, thenetwork102 is depicted as the Internet to whichserver systems104,customer152A,customer152B, andclient systems1202A,1202B . . .1202C are connected. These systems are connected throughcommunication links114,119A,119B,1204A,1204B . . .1204C, respectively. As noted above, these communication links may include any of a wide variety of devices and/or communication techniques for allowing a system to interface with other connected systems.
• As shown inFIG. 12A, and as discussed above, thecustomers152A and152B may desire to send information or projects, such as advertisements (ADV)1206A and1206B and/or projects (PROJ)1208A and1208B, to groups of client systems that have particular or selected capabilities. The number of different groups of client systems is as varied as the capability and user data available for those client systems. The client systems1202A represent client systems that include a first set (Set1) of desired attributes. Theclient systems1202B represent client systems that include a second set (Set2) of desired attributes. And theclient systems1202C represent client systems that include a Nth set (Set N) of desired attributes. Once attributes are selected, the client systems with those attributes may be accessed as desired bycustomers152A20 and152B. For example,customer152A may send its advertisement toclient systems1202B.Customer152B may send its advertisement to client systems1202A. Theproject1208A fromcustomer152A may be processed byclient systems1202C. And theproject1208B fromcustomer152B may be processed byclient systems1202B. It is noted, therefore, that any combination of desired attributes, such as device capabilities and user characteristics, may be identified and utilized to satisfy customer objectives, whether those objectives be advertising, project processing, or some other desired objective.
• FIG. 12B is a block flow diagram for client system attribute selection, according to one or more embodiments. In the embodiment shown,process1250 begins with the customer selecting desired attributes inblock1252. Next, client systems with selected attributes are accessed inblock1254. And, then in block1256, the customer objective, such as advertising or project, is processed by the client system. Control of thisprocess1250 may be provided by theserver systems104, if desired, such that the customer interfaces with theserver systems104 to select device attributes and then theservers systems104 access the client systems. Alternatively, theserver systems104 may simply provide the customer with a list of contact information (e.g., IP addresses) for the client systems, so that the customer may directly access the client system, for example, in providing advertisements to the users of the client systems. It is further noted that other control techniques may also be used to identify and access client systems with particular desired device capabilities, user characteristics, or other device attributes, according to the client system attribute selection method of one or more embodiments.
• FIG. 7A is a block diagram for anetwork102 according to one or more embodiments, includingexample network sites106A and106B on which site testing is to be conducted, such as load testing and/or quality-of-service (QoS) testing.FIG. 7A is similar toFIG. 1A except thatother systems106 inFIG. 1A has been represented in the embodiment ofFIG. 7A withnetwork sites106A and106B.Communication line116A between thenetwork102 and thenetwork site106A represents a interaction by oneclient system108,110 and112. Communication lines116B,116C and116D represent interactions by more than oneclient system108,110 and112.
• Site testing is typically desired to determine how a site or connected service performs under any desired set of test circumstances. With the distributed processing system of one or more embodiments, site performance testing may be conducted using any number ofreal client systems108,110 and112, rather than simulated activity that is currently available. Several tests that are commonly desired are site load tests and quality of service (QOS) tests. Quality of service (QOS) testing refers to testing a user's experience accessing a network site under normal usability situations. Load testing refers to testing what a particular network site's infrastructure can handle in user interactions. An extreme version of load testing is a denial-of-service attack, where a system or group of systems intentionally attempt to overload and shut-down a network site. Advantageously, one or more embodiments will have actual systems testing network web sites, as opposed to simulated tests for which others in the industry are capable.
• Network site106B and the multiple interactions represented bycommunication lines116A,116B and116C are intended to represent a load testing environment.Network site106A and thesingle interaction116A is indicative of a user interaction or QOS testing environment. It is noted that load testing, QOS tests, and any other site testing may be conducted with any number of interactions from client systems desired, and the timing of those interactions may be manipulated and controlled to achieve any desired testing parameters. It is further noted that periodically new load and breakdown statistics will be provided for capacity planning.
• FIG. 7B is a functional block diagram for a site-testing operation700 according to one or more embodiments. Initially,client systems108,110 and112 receive workloads that identify testing procedures and parameters in the “clients receive testing workload”block702. Followingline714, theclient systems108,110, and112 access the site being tested and perform the testing in block “clients interact with other systems”block704. Next, followinglines716 and718, theclient systems108,110 and112 complete the site testing workload tasks, get the results ready for transmission, and send those results back to thesystem server104 in “clients complete testing workload”block706 and “site testing results sent to server system”block708. Control passes alongline720 to “site testing results compiled for use”block510 where the server system formats and/or compiles the results for use by the network site. For example, the site testing results may be utilized determining modifications that are to be made to the network site to handle peek volume activities.
• FIG. 8 is a block diagram for a distributedprocessing system800 for a data back-up system application, according to one or more embodiments. As stated above with respect toFIG. 1A, thenetwork102 may be a wide variety of networks, including an intranet network. Intranet networks, such as internal networks set up by corporations, are particularly suited for this application because the systems holding the data being backed-up would be owned by the same entity owning other systems with excess data storage capabilities. In this way, security would not be as great of an issue and the client system types could be better controlled. It is noted, however, that this data back-up application would be equally applicable to other networks, such as for computer systems connected through the Internet.
• Referring back toFIG. 8,client systems108,110 . . .112 are shown each having backup data blocks804,806 . . .808.Customer systems152 is shown as havingdata802, which is desired to be backed-up with the distributed back-upsystem800. Theserver systems104 manage the flow of data from thedata802 and the client systems that have extra storage space represented by back-up data blocks804,806 . . .808. In operation, theserver system104 identifies client system storage capabilities. With this information, theserver systems104 can receive data for back-up from any system on thenetwork102. It is noted, and as indicated with respect toFIG. 1A, theclient systems108,110 . . .112 and thecustomer systems152 may communicate directly with each other in peer-to-peer type communications.
• Theservers systems104 may also manage the storage and transfer of data so that the data will be readily retrievable once backed-up and stored on theclient systems108,110 . . .112. If desired, a summary index or directory of the backed-up data may be stored centrally on theserver systems104, or may be stored remotely on theclient systems108,110 . . .112. It is also noted that theserver systems104 may also distribute data back-up workloads so that each portion of thedata802 is stored redundantly on at least two of theclient systems108,110 . . .112. This redundancy provides added security should any one or more client systems suddenly cease to be operational.
• Looking now toFIG. 9, a block diagram is depicted of an alternative representation of an interconnection fabric for a distributedprocessing system environment100, according to one or more embodiments. In this diagram and as described above, the network environment may be the Internet, an internal company intranet, a local area network (LAN), a wide area network (WAN), a wireless network, a home network, or any other system that connects together multiple systems and devices. In addition, the server systems and clients systems may be interconnected by a variety of possible connection interfaces, for example, Ethernet connections, wireless connections, ISDN connections, DSL connections, modem dial-up connections, cable modem connections, direct T1 or T3 connections, fiber optic connections, routers, portal computers, as well as any other network or communication connection. It is noted, therefore, as discussed with respect to other embodiments such as the embodiment ofFIG. 1A, that systems may be coupled into an interconnected fabric in any of a variety of ways and communications can potentially occur directly or indirectly between any of the systems coupled into the fabric, as would be understood by those of skill in the art.
• Within this environment, as depicted inFIG. 9,server systems104 are interconnected with any number of client systems, for example,client systems108A,108B,108C,108D,108E,108F,108G,108H,108I,108J,108K, and108L. In addition, these client systems may also includeidle client systems902A,902 B, and902C, as discussed further below. Furthermore, these client systems may include client system904A with a component A,client system904B with a component B, and client system904C with a component C. It is also noted that the interconnection fabric may include any number of devices that are not client systems, in that they themselves are not providing components or processing capabilities for the distributed processing 20 system of one or more embodiments. Nevertheless, these devices may be considered part of the system because they may relay, interpret, process or otherwise transmit or receive information from or to client systems that are part of the distributed processing system.
• Aggregation of component level resources, according to one or more embodiments, will now be discussed. As described above, the capabilities of client systems are determined for purposes of allocating, scheduling and managing distributed processing workloads. In other words, each of the client systems may be made up of many individual subsystems with various capabilities. In some cases, it may occur that particular components on different machines may provide added value if combined or aggregated. Thus, utilizing subsystem or component level resources from a heterogeneous group of devices may be the most efficient or otherwise advantageous way of taking advantage of these resources to complete various desired tasks.
• Referring now more particularly toFIG. 9, theclient systems904A,904B, and904C may have component A, component B and component C, respectively, that are better utilized in combination. For example, client system904A may have a fast processor, a high-speed network connection, but little available storage space.Client system904B may have large amounts of available free storage space but little processing power. Client system904C may also have a fast processor, but relatively little available storage space. In this example, a workload that requires both a large storage capacity and a fast processor may be efficiently completed by dedicating component level resources to various parts of the workload from different machines. Thus, the workload may be managed by having client systems904A and904C processing data stored on and transmitted fromclient system904B. Once clients systems904A and904C process data, this resulting data may then be transmitted back toclient system904B for aggregation and eventual transmission back to theserver systems104. Theclient system904B, therefore, essentially acts as a server for a workload subset, sending out portions of a subset workload, receiving back the processed data, and aggregating the data to build a completed workload subset.
• It is noted that any number of different components from different client systems may be aggregated, as desired. For example, for wireless devices, DSP processing and storage components could be aggregated with components from other client systems. For display devices, graphics rendering power could be aggregated. For relatively dumb machines, such as connected household appliances, vending machines, etc., slow-speed processing components could be aggregated. In short, an appropriate workload may include instructions to numerous client systems that will enable collaboration and aggregation of component level resources. Such instructions may include things, such as, where to receive input, where to send output, and ultimately which client systems return final results.
• It is further noted that the control instructions may be de-centralized as well. In other words, as indicated above, client systems may communicate directly with each other, for example, in a peer-to-peer fashion. In this way, workload communications may occur directly between client systems, and workload control and management may occur through the client system agents located on client systems.
• Still referring toFIG. 9, idle system determination will now be discussed. As stated above, client system capabilities are determined and utilized within the distributed processing system of one or more embodiments. The more time any particular client system is idle, the more processing it is arguably able to accomplish, and the more incentives it is likely to receive. In other words, the client system capabilities may be utilized more often and more intensely if the client system is idle more frequently. As such, it is advantageous to identify idle client systems and allocate them to more processor and time sensitive tasks. By identifying these idle client systems, resources available on the network at any given time may be more fully utilized, and otherwise idle resources may be utilized for highly intensive, real-time activities that would otherwise require dedicated devices. Examples of such real-time activities include data caching, indexing, etc. InFIG. 9, idle client systems are designated as902A,902B, and902C.
• Identifying idle resources may be determined in any of a variety of ways. It is possible, for example, to simply look at whether a machine is not being used or has low processor utilization at any given time. This simple determination, however, may not yield an accurate picture of how idle a client system may or may not be over a given time period. More particularly, discovery methods may be implemented to identify the activity of a variety of client system components and subsystems. For example, subsystems may be monitored, such as network activity, device output activity, user input, processing activity, executing task monitoring, or mode of operation parameters (e.g., mobile or power management modes, stationary or powered mode). In addition, any number of other device vectors may be monitored or analyzed to determine the true usage and idleness of a client system.
• The following TABLE 2 provides a list of idleness vectors or factors that may be utilized in determining the level of device usage or idleness. In particular, TABLE 2 provides two primary categories of activities to monitor or analyze for determination of how idle a client system may or may not be. These activities are user activity and device activity. By monitoring, analyzing, and tracking these client system elements and activities over time, a better determination of device usage and idleness may be made. It is noted that the list provided in TABLE 2 is an example list, and any number of categories, vectors or factors may be identified and utilized, as desired, according to one or more embodiments.

• TABLE 2
  Example Client Idleness Vectors orFactors

  1	User Activity (e.g., monitor input	a.	keyboard input
  	activities, monitor output activities,
  	monitor time elapsed since last input
  	event and between input events, etc.)

  	b.	mouse input	c.	microphone/voice input
  	d.	tablet input	e.	pen input
  	f.	touch screen input	g.	joystick input
  	h.	gamepad input	i.	video output
  	j.	printer output	k.	any other user activity that could be
  				utilized to classify if a device is idle

  2	Device Activity (e.g., monitor	a.	power state (e.g., time since last
  	utilization levels, monitor time elapsed		power state change event)
  	since last device activity, monitor time
  	between changes in device utilization
  	levels, etc.)

  	b.	mobility state (e.g., time since	c.	screen saver activity or trigger (e.g.,
  		device last in mobile state)		time elapsed since screensaver
  				activity or trigger)
  	d.	screen output (e.g., time elapsed	e.	network or communication packets
  		since last screen output, paint event		sent or received (e.g., time elapsed
  		or pixel change)		since last network or
  				communications activity)
  	f.	storage device activity (e.g., time	g.	processor, DSP, microcontroller,
  		elapsed since last storage device		embedded device, or other
  		activity, such as hard drives, flash		processor activity (e.g., time elapsed
  		memory cards, removable drives, CD		since last processor activity)
  		drives, DVD drives, etc.)
  	h.	processor, DSP, microcontroller,	i.	tasks or processes executing (e.g.,
  		embedded device, or other processing		time elapsed since change in number
  		device utilization (e.g., change in		of tasks or processes executing)
  		utilization levels)
  	j.	task or process device utilization	k.	any other device activity that could
  		(e.g., time since change in task or		be used to classify if a device is idle
  		process device utilization)

• As a further example of the usefulness of this determination, reference is made back toFIG. 9.Server systems104 may have, for example, a large, intensive task that it would like to place on these idle devices. After using a number of the vectors in TABLE 2 to determine the utilization level for client systems, theserver systems104 determines thatclient systems902A,902B, and902C are idle and capable of handling significant time sensitive processing tasks. For example,idle client systems902A,902B and902C may be personal computers that can act as a local interne cache for other connected devices, such as some of the other client systems depicted inFIG. 9, that are interested in a data type that benefits from a local network cache.
• Thus, data or content may be transmitted from a remote network site to theidle machines902A,902B, and902C. Theseidle devices902A,902B, and902C may then re-transmit this same data or content to other connected devices also interested in the data or content. One example for such network caching is Internet video or multimedia broadcast events that are desired to be viewed or received by a very large number of geographically close connected devices• at about the same time. In order to meet the demand of these connected devices, web sites broadcasting an event have to be able to handle a huge increase in network traffic over a short period of time. By locally caching the transmission to idle client systems, a web site can reduce the direct demand on its own resources. This is so because other connected devices may receive a re-transmitted broadcast, although delayed, from the idle client system. It is noted that according to one or more embodimentsidle client systems902A,902B and902C may work independently or in combination. Even though idle client systems are suited for providing the caching function, it is also noted that that network caching may be accomplished using one or more client systems regardless of their respective levels of idleness.
• FIG. 10 is a more detailed block diagram for aclient system agent270 installed on a client system, according to one or more embodiments. This diagram includes asecurity subsystem1010, acapabilities subsystem1006, aworkload processor1004, auser interface1002, and a project management andagent control subsystem1008. The various components and subsystems may communicate with each other, for example, throughlines1012,1014,1016,1018, and1020. Externally, theclient system agent270 may communicate through itssecurity subsystem1010 with the other components within the client system and ultimately to other devices connected into the network fabric. It is noted that configuration of the client system agent and its operation, both internal and external, may be selected and designed, as desired. As depicted, thecapabilities subsystem1006 includes anidle system monitor1022, as described above, that monitors and analyzes user and device activities associated with the client system to determine the level of activity or idleness for the client system. The information determined by this idle system monitor1022 may then be communicated externally, for example, through thesecurity subsystem1010 to theserver systems104. Theserver systems104 may then store and analyze system idleness data from across the distributed processing system. This idleness data may become part of the capabilities database that is utilized to allocate and manage workloads and processing system resources.
• Still referring toFIG. 10, theworkload processor1004 includes a machineentry generation subsystem1024. As described above, theworkload processor1004 may send completed workloads back toserver systems104 to generate sweepstakes entries for the host client system. In this way, when the incentive is a sweepstakes, the client system may generate entries by completing workloads. The machineentry generation subsystem1024 refers to this entry generation through workload completion. As discussed above, the workload processed to generate entries may be a project workload, an entry workload, or any other workload, as desired.
• FIGS. 11A and 11B provide more detailed flow diagrams of process embodiments for machine generated sweepstakes entries through processing of entry workloads, according to one or more embodiments.
• Looking first toFIG. 11A, an entryworkload process flow1100 is depicted that provides machine generated sweepstakes entries. Process moves fromstart block1102 to block1104 in which entry workloads are loaded on client systems. Next, process flows to block1106 which represents a periodic timer or other timing control for entry workload processing. After this timing control, the client system executes or processes the entry workload inblock1108. Inblock1110, a sweepstakes entry is thereby generated and returned to theserver system104 based upon the completion of this entry workload. Process control then may proceed back to theperiodic timing block1106, where timing control determines when the entry workload is next processed. The completed workload represents the machine generated sweepstakes entry.
• FIG. 11B is an alternative entryworkload process flow1150. Theprocess flow1150 is similar to theprocess flow1100 except that the entry workload is sent to the client system each time it is to be run. Process starts inblock1102 and passes to theperiodic timer block1106, in which the process is controlled. For example,server systems104 may determine when it is desirable for the client systems to receive and process an entry workload. Inblock1104, the entry workload is sent to the client systems. As withFIG. 11A, the client systems then execute the entry workload inblock1108, and an entry is generated and returned to theremote server systems104 inblock1110. The process then proceeds back to theperiodic timer1106 until it is determined that another entry workload should be processed. The primary difference betweenprocess1150 andprocess1100 is thatprocess1150 is depicting an entry workload that is transmitted to the client system each time it is to be run. One example utilizing theprocess1150 or theprocess1100 is forservers systems104 to query the client systems for entry workload processing at regular time intervals. If a distributed device returns a completed entry workload back within a selected period of time from the distribution of the entry workload, the server system may conclude that the distributed device should receive an entry because the distributed device is providing resources to the distributed processing system. In this way, theserver systems104 may determine at regular intervals whether a given client system is working on project workloads for the distributed processing system. Alternatively, the client system agent may locally control the workload processing and may, for example, cause the client system to process and generate entries at regular time intervals. It is noted that non-regular and varying time intervals may also be utilized and that combinations of remote and local control may also be utilized, as desired.
• The timing of when a client system processes the entry workload, therefore, may be determined locally by the client system agent or remotely, for example, through commands sent by theserver systems104. In addition, periodic timing control may also be accomplished through various combinations of control routines residing locally and remotely. It is further noted that any number of different variations may be utilized to provide machine generated entries to a sweepstakes, according to one or more embodiments. Thus, a client system may generate sweepstakes entries in any of a variety of ways and still have machine generated sweepstakes entries, according to one or more embodiments.
• FIGS. 13A and 13B describe adata conversion application1300 for a massively parallel distributed network according to one or more embodiments. In particular,FIG. 13A is a block diagram of a distributed processing system that provides data conversion services, according to one or more embodiments. AndFIG. 13B is a block flow diagram for data conversion services within a distributed processing system, according to one or more embodiments.
• Converting file types, web pages, graphics images, etc., between device types can be a highly intensive processing task. Example devices that often use converted data are wireless devices, such as pagers and cell phones, that request Internet web page information from their respective device servers. The device server, instead of incurring the overhead of reformatting the requested data for the wireless devices, may instead distribute the requested page or data address, the device type information of the requesting device, and return address for the reformatted data. According to one or more embodiments, the data conversion, translation or processing may be performed by a client system of the distributed processing system of one or more embodiments. The resulting data may then be returned or provided to the original requesting device. In addition to data formatting for cell phones, language conversion, text translation and media translation services, or any other desired data conversion can also be hosted for a customer through the distributed processing system of one or more embodiments.
• It is noted that the data conversion operation contemplated by one or more embodiments is not limited to any particular requesting device, any particular service provider, any particular type of data to be processed, any particular type of resulting processed data, or any particular data source. Thus, the data processed may include voice, text, application, image, source code, or any other data type or combination of data types, and the resulting processed data may also include voice, text, application, image, or any other data type or combination of data types. According to one or more embodiments, the distributed processing system is utilized to process any data that is desired by a requesting device and that must be converted or processed before being provided to the requesting device. For example, an end-user devices connected to the Internet, such as personal computers, may sign up for data conversion services through the server system so that the end-user device may request data conversion of any desired data, file, web site content, etc. Language translations and data formatting for connected wireless are just two examples of such applications for one or more embodiments.
• Looking now to the embodiment ofFIG. 13A, thenetwork102 is depicted as the Internet, and the requesting device is one ormore wireless devices1306 connected to theInternet102 throughcommunication links1308 and to the wirelessdevice server systems1304 throughcommunication link1309. The data to be converted, translated or otherwise processed is represented byblock1302 and may be, for example, content from an Internet web site that is connected to the Internet throughcommunication link1312. Also, as shown inFIG. 13A, a massively parallel distributed network (MPDN)server104 is connected to theInternet102 throughcommunication link114. The wirelessdevice server systems1304, or any other connected system that desires to off-load data conversion processing requirements (e.g., web site content servers), are connected to theInternet102 throughcommunication links1310 and to theMPDN server104 throughcommunication links1311. Any number ofclient systems108,110,112 may also be connected to theInternet102, throughcommunications links118,120 . . .122, respectively. As also stated above, any of the connected devices may communicate with each other in any of a wide variety of communication techniques (e.g., wireless, electrical, digital, analog, light-based, etc.) and protocols (e.g., static or dynamic EP addresses), and through any number of other devices, as would be understood by one of skill in the art.
• In the application contemplated byFIG. 13A, thewireless devices1306 at times request data, for example, images or text from a web site, that must be converted, translated or otherwise processed by wirelessdevice server systems1304 before it can be transmitted to, and displayed on, a requesting wireless device. Instead of converting the information, the wirelessdevice servers systems1304 may request that theMPDN server104 accomplish the data conversion or translation. Thedevice server systems1304 may then provide to theMPDN server104 any pertinent information, such as information concerning the requesting device, the nature of the data requested, and the processing utilized for the data. TheMPDN server104 may then utilize one or more of theclient systems108,110 . . .112 to process the data fromblock1302 for transmission to the requesting device. In this way, the wirelessdevice server systems1304 may off-load burdensome and process-intensive conversion tasks to the distributed processing system of one or more embodiments.
• It is noted the transmission of processed data to the requestingwireless device1306 may occur in a variety of ways. For example, the processed data may be transmitted from aclient system108 to theserver104, then to thewireless device server1304 and finally to thewireless devices1306. Alternatively, the processed data may be transmitted from a client system to thewireless device server1304, and then to thewireless devices1306. Still further, the processed data may be transmitted directly from a client system to the wireless devices.
• FIG. 13B provides a basic flow diagram for an embodiment of adata conversion process1350 according to one or more embodiments. Inblock1352, a device, such aswireless devices1306, requests unconverted, non-translated or non-processed data. Inblock1354, a server for the device, such as wirelessdevice server systems1304, processes the data request and contacts theMPDN server104. In addition, the content provider or server for the requested data, such as a web site content server, may contact theMPDN server104. The wirelessdevice server systems1304 provide all pertinent information to theMPDN server104, such as the type of calling device, its identification, the relevant data requested, and the conversion to take place. TheMPDN server104 then distributes the data and information concerning the requesting device to one or more client systems, such asclient systems108,110 . . .112, inblock1356. The one or more client systems then convert, translate or otherwise process the data inblock1358. The converted, translated or processed data is then provided to the requesting device inblock1360. Again, in this way, the device servers may provide a wide range of information without having to provide itself the processing power to accomplish the conversion, translation or processing that is required to transmit or display the data on a requesting device.
• As shown inFIG. 13B, the device server or thecontent server1304 may communicate data and other pertinent information for a conversion directly to the client systems. For example, theMPDN server104 may provide access to a group of client systems for data conversion purposes for given periods time (e.g., monthly client group allocations), or may provide identities of groups of client systems that may be used at the time a conversion is to be used. Once the identity and allocation of client systems to a particular device server or content server is made, the device server or content server may communicate directly with the client systems. In addition, the device server or content server may provide directly to a requesting device the identity of the one or more client systems accomplishing the data conversion. As shown inFIG. 13B, the requesting device, therefore, may communicate directly with the client system or systems to provide pertinent information concerning the data conversion requested. The client system may then, for example, directly download the desired content and perform the desired data conversion. It is further noted that in addition to the embodiments described above with respect toFIGS. 13A and 13B, other methods for requesting, processing and providing data to and from the requesting device may be implemented with distributed processing system of one or more embodiments, such as caching processed data for later transmission.
• FIGS. 14A and 14B depict example block diagrams of file distribution and data sharing through the network fabric, according to one or more embodiments. In particular,FIG. 14A depicts an Internet datafile distribution system1400 that relies upon client systems to provide local data distribution.FIG. 14B depicts a datafile distribution system1450 that allows for data sharing and rapid transmission of a project or data files through the distributed processing system.
• Looking now toFIG. 14A, a block diagram is depicted of a distributedprocessing system1400 that provides data transmission caching or other local distribution, according to one or more embodiments. In the embodiment ofFIG. 14A,server systems104 are connected throughcommunication link114 to the Internet backbone1402. The Internet backbone1402 represents the very high speed connections that carry data long distances, for example, T3 or fiber optic lines that carry Internet data across the United States. Aweb site1404 is connected to the Internet backbone1402 throughcommunication link1406, which represents a geographically local connection. Theconnection block1410 represents a geographically remote communications link, such as a POP server, head-end machine, telephone line central office, cell site, etc. Thiscommunications block1410 is connected to the Internet backbone1402 with acommunications link1408, which also represents a geographically local connection. A variety of client devices andnon-client devices1412A,1412B,1412C,1412D,1412E and1412F may be connected below theconnection block1410. It is noted thatinterface1414 represents, for example, a secondary network on whichclient devices1412D,1412E and1412F are connected, such as a home network.
• In the embodiment shown inFIG. 14A,web site1404 may provide content that is in high demand, over a short period of time. An example of such an event is a live Internet multimedia broadcast. For such an event, there may be a huge influx of devices trying to download the content from theweb site1404 over a short period of time. Theweb site1404 may be unable to meet this extremely large demand, requiring theweb site1404 to shut down.
• According to one or more embodiments, theweb site1404 may off-load some or all of its data handling requirements by using the distributed processing system for data caching. Theweb site1404 may contactserver systems104 and request data caching services. Theserver systems104 may then identify a local machine, such asclient device1412E, to act as a local distributor of the content forweb site1404. For example, one or more idle client devices that have been identified, as discussed above, may be utilized as localdistributor client device1412E. The localdistributor client device1412E may first download the content and pass it on to other client andnon-client devices1412B,1412C, and1412D throughcommunication links1416A,1416B, and1416C. It is noted that this caching will be aided if the client and non-client devices receiving the cached data are relatively short communication hops from localdistributor client device1412E.
• This data or network caching allows data to be streamed to an end user level device, which may then pass the data on to other end user devices. Thus, the downstream communications may be limited, thereby taking the distribution burden off of the web site. For example,web site1404 may have a large streaming video or multimedia file that is experiencing a heavy load from a given set of network devices. This data file may be cached by a machine, such asclient device1412E, that is below from acommunication link1410. Then, other devices that are also below thiscommunication link1410 may download the streaming video data from theclient device1412E. This caching can eliminate repeatedly sending the same data through the same communication links to requesting devices that are located below common communication links. It is noted that the file and data distribution possibilities for this peer file access, caching and data transmission, according to one or more embodiments, are wide and varied and should not be seen as limited to the embodiment shown inFIG. 14A.
• FIG. 14B is a block diagram of a distributedprocessing system1450 that provides data distribution and data sharing, according to one or more embodiments. As withFIG. 9,FIG. 14B depicts an alternative view of a network fabric that may interconnect any of a wide variety of devices. In the embodiment shown inFIG. 14B,server systems104 are interconnected with any number ofclient systems108A,108B,108C,108D,108E,108F,108G, and108H. Each of the connecting interconnects represents any of a wide variety of communication links that may exist between devices in the network fabric of one or more embodiments. Each of theclient systems108A,108B,108C,108D,108E,108F,108G, and108H include shared data (SD) according to one or more embodiments. Within this interconnected fabric,block1452 represents data or project information that is desired to be distributed. The SD blocks within each client system facilitates the distribution of this data or project information.
• A client agent, as discussed above, installed on theclient systems108A,108B,108C,108D,108E,108F,108G, and108H includes functionality that facilitates a number of services with respect to data transmission and sharing. First, the client agent provides a protected data storage area accessible to outside devices, which is represented by the SD block within each client system inFIG. 14B. This special storage space protects the device from outside devices accessing other storage areas on the device while allowing data to be shared and accessed by other devices and simultaneously used by the local client agent.
• These shared data (SD) blocks provide mechanisms that enable a wide variety of possible interactions among theclient systems108A,108B,108C,108D,108E,108F,108G, and108H. For example, the data sharing mechanism may provide a space for a cache of other device addresses attached to the network for both communication purposes as well as security purposes. The mechanism may also provide a simple indexing system that is automatically re-indexed when content is added or removed from the storage area. This indexing system may provide a mechanism for other client agents to perform discovery on the local client information and vice versa. Through information stored within this shared data, the distributed processing system of one or more embodiments facilitates many distributed file system applications such as distributed resume posting, distributed caching, distributed advertisement serving, etc. In addition to the above, the storage block (SD) within each client system may include an interface for displaying or playing data types (such as images, audio files, video fifes, etc.) stored both locally and/or remotely on other client devices. This would enable simple picture sharing, for example, between remote families connected via the Internet, as part of being a client system within the distributed processing system of one or more embodiments.
• In the embodiment shown inFIG. 14B, data orproject1452 is injected into the fabric through a connection toclient system108C andserver systems104. These connections represent that the information may pass first toservers systems104, or may pass first to a client system, such asclient system108C. It is noted that there are other ways that the data may be injected into the fabric. Once injected, thedata1452 may be transmitted throughout the fabric through any of a wide variety of communications, including client-to-client, server-to-client, client-to-server, client-to-non-client, non-client-to-client communications, and/or non-client-to-non-client communications. These communications may be based upon a variety of mechanisms, such as polling mechanisms and pre-assigned firewall ports. This technique provides a vehicle that facilitates the distribution of information to a large number of devices in a short period of time.
• Applications for this data distribution are wide a varied. For example, a file that is time sensitive may be propagated to a large number of client devices, non-client devices, servers, or other connected devices, in a short amount of time. This transmission may occur quickly and efficiently once the information is injected into the distributed processing system of one or more embodiments. Example time sensitive data files are anti-virus signature files, which when distributed through the distributed processing system of one or more embodiments, may be transmitted through the network fabric faster than a new virus may normally proliferate.
• Another application for rapid propagation of files is utilizing this technique for propagation of workloads. One example is distributed resume or job searching. In such a system, participating job seekers and participating employers may rapidly search for one another. A job seeker may inject a job request or search into the fabric that is then routed by each successive device to other devices without control from theserver systems104. Similarly, an employer may inject candidate criteria into the fabric that is then routed to successive devices. The result is an extremely fast search and identification of employers and candidates.
• FIG. 15 is a block diagram of an alternative representation for a distributedprocessing system100, according to one or more embodiments.Server systems104,database systems1546, andweb interface1554 are coupled together throughcommunication links1540,1542, and1544. Theweb interface1554 includes client'ssubsystem1548,task developer subsystem1550, and advertiser'ssubsystem1552, and may include other subsystems as desired. Thedatabase systems1546 include workload (WL)information308, clientcapability vector information620, and any other stored information as desired. Server systems include various modules and subsystems, including database interface1532,web server1536, task module andwork unit manager1530, client statistics module1534, advertising manager1538, task module version/phase control subsystem1528,sweepstakes engine1524,server control subsystem1526, andcommunication interface1522. It is noted that in the embodiment of a distributedprocessing system100 as depicted in ofFIG. 15, the three primary operations for theserver systems104,database systems1546 andweb interface1554 are directed to managing, processing and providing an interface for client systems, customer tasks, and customer advertising.
• As discussed above, each client system includes a client agent that operates on the client system and manages the workloads and processes of the distributed processing system. As shown inFIG. 15, each of theclient agents270A,270B . . .270C communicates with theserver systems104 throughcommunication links1516,1518 . . .1520, respectively. As discussed above, any number of different techniques and architectures may be utilized to provide these communication links. In the embodiment as shown inFIG. 15 with respect toclient agent270A, each client agent includes a base distributedprocessing system component1506 and a separate project orworkload component1504. As depicted, acommunication interface1508, acore agent module1502, and auser interface1510 make up the base distributedprocessing system component1506. Thetask module1512 and thework unit1514 make up the separate project orworkload component1504. Thetask module1512 operates on top of thecore agent module1502 to provide processing of eachproject work unit1514. It is noted that different or additional modules, subsystems or components may be included within the client agent, as desired. For example, a personal computer screen saver component may be part of the base distributedprocessing system component1506 or the separate project orworkload component1504.
• Also as discussed above, security subsystems and interfaces may be included to provide for secure interactions between the various devices and systems of the distributedprocessing system100. As depicted inFIG. 15, a security subsystem andinterface1560 is interconnected with theserver systems104, thedatabase systems1546, theweb interface1554, and theclient agents270A,270B . . .270C. These interconnections are represented bylines1566,1564,1562, and1568, respectively. The security subsystem andinterface1560 operates to secure the communications and operations of the distributed processing system. This security subsystem andinterface1560 also represents a variety of potential security architectures, techniques and features that may be utilized. This security may provide, for example, authentication of devices when they send and receive transmissions, so that a sending device verifies the authenticity of the receiving device and/or the receiving device verifies the authenticity of the sending device. In addition, this security may provide for encryption of transmissions between the devices and systems of the distributed processing system. The security subsystem andinterface1560 may also be implemented in a variety of ways, including utilizing security subsystems within each device or security measures shared among multiple devices, so that security is provided for all interactions of the devices within the distributed processing system. In this way, for example, security measures may be set in place to make sure that no unauthorized entry is made into the programming or operations of any portion of the distributed processing system including theclient agents270A,270B . . .270C.
• In operation, client systems or end-users may utilize the client'ssubsystem1548 within theweb interface1554 to register, set user preferences, check statistics, check sweepstakes entries, or accomplish any other user interface option made available, as desired. Advertising customers may utilize the advertisers subsystem1552 within theweb interface1554 to register, add or modify banner or other advertisements, set up rules for serving advertisements, check advertising statistics (e.g., click statistics), or accomplish any other advertiser interface option made available, as desired. Customers and their respective task or project developers may utilize thetask developer subsystem1550 to access information withindatabase systems1546 and modules within theserver systems104, such as the version/phase control subsystem1528, the task module andwork unit manager1530, and theworkload information308. Customers may also check project results, add new work units, check defect reports, or accomplish any other customer or developer interface option made available, as desired.
• Advantageously, the customer or developer may provide the details of the project to be processed, including specific program code and algorithms that will process the data, in addition to any data to be processed. In the embodiment shown inFIG. 15, this program code takes the form of atask module1512 within the workload, while the data takes the form ofwork unit1514. These two portions make up the project orworkload component1504 of eachclient agent270. For a given project, thetask module1512 will likely remain relatively constant, except for version updates, patches or phase modifications, while thework unit1514 will likely change each time processing of the data that it represents is completed. The project orworkload component1504 runs in conjunction with the base distributedprocessing system component1506. When a different customer or project is started on a given client system, the project orworkload component1504 will typically be replaced, while the base distributedprocessing system component1506 will likely remain relatively constant, except for version updates, patches or other modifications made for the distributed processing system.
• Information sent from theservers systems104 to theclient agents270A,270B . . .270C may include task modules, data for work units, and advertising information. Information sent from theclient agents270A,270B . . .270C to theserver systems104 may include user information, system information and capabilities, current task module version and phase information, and results. Thedatabase systems1546 may hold any relevant information desired, such as workload information (WL)208 and client capability vectors (CV)620. Examples of information that may be stored include user information, client system information, client platform information, task modules, phase control information, version information, work units, data, results, advertiser information, advertisement content, advertisement purchase information, advertisement rules, or any other pertinent information.
• Now looking toFIGS. 16,17A,17B,18A and18B, an embodiment for security features for the distributed processing of the one or more embodiments will be described.FIG. 16 provides a representation of the distributed processing environment including security subsystems.FIGS. 17A and 17B provide block diagrams of the communication interface between client systems and the server systems. AndFIGS. 18A and 18B provide detailed block diagrams of an embodiment of security measures for the servers systems and the client systems.
• Referring toFIG. 16, anembodiment1600 of a distributed processing system is depicted.Server system104 includes asecurity subsystem354 through which communications to and from theserver systems104 may be made secure.Client systems108A,108B . . .108C andclient systems108D,108E . . .108F represent any number of client systems that may communicate withserver systems104 or with each other. Each of theclient systems108A,108B,108C,108D,108E, and108F include asecurity subsystem272A,272B,272C,272D,272E, and272F, respectively. Theelectronic information1602 represents information that theserver systems104 is to communicate toclient systems108A,108B,108C,108D,108E, and108F in a secure manner, so that no unintended or intercepting recipient may understand or tamper with theelectronic information1602, and so that no third party may insert non-authorized information into the distributedprocessing system1600. Although not shown, it is understood that any one of theclient systems108A,108B,108C,108D,108E, and108F may have electronic information that is to be securely sent to theserver systems104 or to any other of theclient systems108A,108B,108C,108D,108E, and108F.
• Electronic information1602 represents information that is communicated to facilitate the operations of the distributedprocessing system1600. Such information includes the client agents that are downloaded to each client system, the workload applications for any given workload, and any work unit that will be processed by a client system.Electronic information1602 may also be any type of information to be sent or received within the distributed processing system, such as text, images, audio streams, video streams, databases, spreadsheets, PDF files, Shockwave data, Flash data, applications, data files, chat streams, or any other information, data or data streams. In addition, electronic information may be sent byclient systems108A,108B,108C,108D,108E, and108F to theserver systems104 and/or any of the other client systems.
• The Certificate Authority (CA)block1604 within theserver systems104 represents an entity that helps to ensure validity of encryption and decryption codes. For example, within a public/private key encryption environment, a Certificate Authority may help ensure that a public key alleged to be from a particular entity is in fact legitimately from that entity. One third-party entity that performs this CA function on the Internet is VeriSign, Inc. Having a third-party perform the CA function can be advantageous in a transaction or communication between non-trusted entities. For example, the sending entity provides its public key information to the third-party CA, which verifies the information and creates a certificate that includes the sending entity's public key information. This certificate may then be encrypted and signed by the third-party CA. The receiving entity may then obtain the certificate from the third-party CA and decrypt it with the third-party CA's public key. The receiving party will then have the sending party's public key and be fairly secure that it is a legitimate public key from the sending party.
• As shown inFIG. 16, the CA functionality may be part of theserver systems104, such that theserver systems104 act as their own Certificate Authority with respect toclient systems108A,108B,108C,108D,108E, and108F and any other devices that are part of the distributed processing system. A third-party CA may not be used as much in this distributed processing environment because theserver systems104 primarily direct the operations of the distributed processing system. Thus, a third-party entity may or may not provide a CA function. It is noted that CA functionality may be provided only by theservers systems104, only by third-party CAs, or any combination ofserver systems104 and third party CAs, as desired for a particular embodiment. In addition, if desired, no CA functionality could be provided so that secure communications between theserver systems104 and the devices within the distributed processing system were conducted without the use of a Certificate Authority.
• FIG. 17A is a block diagram of anembodiment1700 for a communication interface between aclient system108 and theserver systems104. In thisembodiment1700, the network can be the Internet. As depicted, theclient system108 includes aclient agent270 and anetwork browser1702. Theserver system104 includes a clientagent download site1710, from which theclient system108 may download theclient agent270 throughcommunications1704. Theserver system104 also includesblock1718, which represents a variety of client service functions that may be provided by the web interface for theserver systems104 throughcommunications1706. For example, in a public/private key security environment, aclient system108 may download from block1712 a Certificate Authority (CA) certificate that includes the server public key. In addition, theclient system108 may login to the web page interface for theserver systems104. And theserver systems104 may generate dynamic certificates. Theclient system108 may also send and receive information toapplication server1714 throughcommunications1708, for example, to receive project work units. Finally, as depicted,database systems1546 may send information to and receive information from theblocks1710,1712 and1714 of theserver systems104 throughcommunications1716,1718 and1720. As discussed more above,database systems1546 may include any desired information, for example, aworkload database308 and/or acapability vector database620.
• FIG. 17B is a block diagram for an Internetcommunication protocol structure1750 that may be utilized forcommunications1704,1706, and1708. As depicted inFIG. 17B, three basic application layers are utilized by eachclient system108 and theserver systems104 to communicate with each other. The TCP/IP layer1756 represents a standard Internet communication protocol that allows devices to identify and send information to each other across the Internet, as is well known to those of skill in the art. The secure network layer (SNL)1754, such as the secure socket layer (SSL), represents a protocol that allows devices to confirm the identity of servers and the other devices with whom they communicate, as long as those servers or other devices utilize similar protocols. Theapplication security level1752 represents other desired security or communication protocols that are implemented by programs running on theclient system108 and/or theserver systems104.
• In operation, theserver system104 may secure the download of theclient agent270 to theclient system108 by requiring that theclient system108 download theclient agent270 from the clientagent download site1710. As part of the server authentication sequence, thedownload site1710 will send back an identifier to assure users that they are indeed connected to theproper server systems104. This identifier may be, for example, a CA certificate, but may be any other identifier, as desired. Because it is desirable to have the client agent running on as many distributed devices as possible for the distributed processing system of one or more embodiments, user authentication may not be required to download theclient agent270 from thedownload site1710.
• Once aclient system108 has downloaded and installed theclient agent270, theclient system108 will communicate with theapplication server1714 to begin working within the distributed processing system. For these communications, server and client authentication may be required to help ensure security. To accomplish this authentication, for example, two-way authentication may be utilized. To provide a public/private key combination for the client agent20270, eachclient agent270 that is downloaded by aclient system108 may have embedded within its code a default identifier and a default public/private key pair. Thus, theserver systems104 may use secure network protocols (such as SSL or similar schemes) to authenticate eachclient system108, and eachclient system108 may use compatible protocols to authenticate each server application with which it communicates. These applications, for example, may include the functionality provided byblocks1712 and1714, and, therefore, thecommunications1706 and1708 would utilize authentication.
• As an alternative to embedding a public/private key combination and associated identifiers or certificates into theclient agent270, the public/private key pairs may be dynamically generated in block1712. For example, at start-up, at reboot or at some desired time or event, theclient system108 may generate a new public/private key pair. When theclient system108 next communicates with theserver systems104, theclient system108 request a certificate from theserver systems104. Theserver systems104 may then act as a Certificate Authority (CA) and provide a CA certificate to theclient system108. This dynamic certificate generation, therefore, allows for added security by allowing eachclient system108 to have its own public/private key pair for secure network protocol communications and by having this key pair change at some desired recurring event for theclient system108, such as reboot.
• Theclient system108 may initiate communications with theserver systems104 by logging on to the authentication server, which may be part of block1712. The user may be prompted to enter a valid e-mail address and/or password, if already registered, or may be asked to register if the e-mail address and/or password are not recognized. Once registration is completed, a password may be e-mailed back to the user to provide validation of the user. If authentication is successful when a user logs into theserver systems104, theserver systems104 may provide a host-ID, a user-ID, and a session key for any given communication session.
• It is also desirable that once a user has successfully registered, the user may install theclient agent270 on any number of other host or user systems without interacting with that systems network browser, other than to set host-specific preferences. For example, when downloaded, theclient agent270 may take the form of a self-extracting program that installs appropriate files to theclient system108, including the proper host and user identifications. In addition, to help ensure proper identification, the session keys may be exchanged each time theclient system108 communicates with theserver systems104. For example, theclient system108 may communicate its current session key to theserver systems104 each time it communicates with theserver systems104. Theserver systems104 will then send a new session key for theclient system108 to utilize for the next session. In this way, stale identification information may be reduced. In addition to this security feature, communications may also be encrypted and decrypted with various encryption techniques, as desired.
• Referring now toFIGS. 18A and 18B, one embodiment will be discussed for a security model utilizing public/private key encryption. This security model utilizes a third-party CA to provide a CA certificate for theserver systems104.
• FIG. 18A is a block diagram of anembodiment1800 for security procedures implemented byserver systems104.Electronic information1602 is communicated to aclient system108. Thiselectronic information1602 travels through four different paths that provide security information. One path begins with theelectronic information1602 being encrypted with the server private key inblock1802. Then, in block1830, the encrypted information is sent to client systems. This encrypted information is represented byarrow1826. A second path flows fromblock1802 to block1804 where a hash value is generated for the encrypted electronic information. It is noted that a hash value is a unique value that may be generated for any given electronic file based upon the contents of that file and the algorithm used to calculate the unique value. There are any number of algorithms that may be used to calculate a hash value, as would be understood by one of skill in the art. Proceeding down the second path to block1806, the hash value generated on the server side for the encrypted electronic information (i.e., the information sent to the client system in1830 via1826) is compared with ahash value1822 from theclient system108. Thishash value1822 represents the client system's calculation of the hash value for the encrypted electronic information that theclient system108 received from theserver system104. If no tampering has occurred and the data was transmitted accurately, the client system hash value should match the server hash value. Inblock1808, theserver systems104 provide an indication of the result of the hash check evaluation back to theclient system108. This pass/fail determination is indicated byarrow1824.
• A third path begins withblock1810 where a hash value is calculated for non-encryptedelectronic information1602. This hash value is then encrypted inblock1816 with the server private key. Next, this encrypted hash value is sent to theclient system108 inblock1818. Thearrow1821 represents the encrypted hash value for the non-encrypted electronic information. A fourth path, and the last depicted in theembodiment1800 ofFIG. 18A, flows fromblock1810 to block1812, where the hash value is partitioned into N different portions. These N different portions are designated for Ndifferent client systems108, as well as anyclient systems108 receiving a redundant distribution of any one of the N different portions. Inblock1814, the N different hash value portions are encrypted with the server private key. Next, the N different encrypted hash value portions are sent inblock1820 to Ndifferent client systems108, as well as being sent toclient systems108 receiving redundant distributions of the hash value portions. Thearrows1828 represent the distribution of the N different hash value portions. It is noted that redundant distribution of the N hash value portions is desirable because, as discussed below with respect toFIG. 18B, when the hash value is reconstructed by aclient system108, it is desirable to have multiple sources for each portion in case one of the receiving client systems is not available at any given time.
• Looking now toFIG. 18B, the corresponding security procedures implemented byclient system108 are discussed. Initially, atblock1854, theclient system108 receivesCA certificate1852 containing the server public key and the server identity. It is again noted that other unique identifiers may be utilized instead of CA certificates, as described above. If a CA certificate is utilized, this CA certificate may be provided from a third-party Certificate Authority (CA) or from theserver systems104 or any other desired source. Inblock1856, theclient system108 verifies the accuracy of the CA certificate using the CA's public key. If this verification is not successful, theclient system108 may wait some period of time before retrying. In addition, the time period may be a random period of time. In addition, as discussed with respect toFIGS. 17A and 17B, theclient system108 will login to theserver systems104. If this authentication is not successful in this login, the client system will notify the user of the system and theserver systems104, and then wait for some period of time or a random amount of time before attempting to re-verify.
• Inblock1862, theclient system108 receives theencrypted information1826. Next, theclient system108 creates a hash value for the encrypted information inblock1864. This hash value can be calculated using the same algorithm utilized by theserver systems104 in generating the hash value for the encrypted information inblock1804 ofFIG. 18A. Once theclient system108 has calculated the hash value for the encrypted information, thishash value1822 is sent to the server systems inblock1866. As discussed above, a pass/fail response1824 is sent back by theserver systems104. This hash check evaluation is received in block1868. If the check was a FAIL, flow passes to block1870 where theclient system108 sends out a notice to theserver systems104 and any other client system to which it is attached that a problem has been encountered. Theclient system108 then ends the current connection with theserver systems104. It is noted that theclient system108 may retry several times before moving onto block1870, and that the reporting scheme may be modified, altered or developed as desired.
• If the hash check evaluation was a PASS, flow passes to block1872 where the electronic information is decrypted with the server public key, which was verified inblock1856. A hash value is then calculated for theelectronic information1874. Again, the hash generation algorithm can be the same as that used by theserver systems104 in creating the hash value inblock1810 ofFIG. 18A. Next, the hash value is sent fromblock1874 to block1886, where it is compared with two other hash value calculations. One of the other hash values comes from a path that begins withblock1858, in which theclient system108 receives theencrypted hash value1821 for the non-encrypted information. Inblock1860, the encrypted hash value is decrypted with the server public key. The hash value is then sent to block1886. The third hash value for block1886 comes from a path that utilizes the N different hash portions sent out by the server systems inblock1820 ofFIG. 18A. Inblock1876, the client system receives a portion1828A of the partitionedhash value1828. In addition to one of the partitioned hash values, it is noted that theserver systems104 will also send information providing the identity and source for the N−1 other hash value portions. In block1878, theclient system108 decrypts the portion1828A with the server public key. Next, in block1880, theclient system108 resolves the identity of the source for the N−1 other portions, which may be N1 other client systems. Inblock1882, theclient system108 obtains the N−1 other portions, and assembles the N partitions into a hash value for the non-encrypted electronic information in block1884. The resulting hash value is then sent to block1886. It is noted, as indicated above, that redundant distribution of the N portions of the partitioned hash value is desirable so that unavailability of one client system will not cause another client system to be unable to reassemble the N different portions.
• Once the three hash values are received in block1886 from three different sources, they are compared to see if they match. If this check is a FAIL, flow moves to block1888, where theclient system108 sends out a notice to theserver systems104 and any other client system to which it is attached that a problem has been encountered. Theclient system108 may also inform the client systems from which it received the N−1 other portions, and theclient system108 may retry the procedures, if desired. In addition, once aclient system108 is notified of a potential problem, theclient system108 may download a special check file from theserver systems104 to make sure that the server systems have not been compromised. If still a FAIL, theclient system108 then ends the current connection with theserver systems104. If the check is a PASS, the electronic information is utilized, as represented byblock1890.
• FIGS. 19 and 20 provide block diagrams for further describing the distributed processing system and environment of one or more embodiments that allows for third parties, such as network service providers, to monetize, or gain revenue from, their respective user bases.
• Looking first atFIG. 19, a block diagram is depicted for a distributedprocessing system100 and anenvironment1900 in which network service providers are enabled to monetize their user bases.Environment1900 includes a distributedprocessing system100, acustomer152, and a third-partynetwork service provider1902. Thecustomer152 represents an entity that has aproject1912 that thecustomer152 would like processed by the distributedcomputing system100. In return for processing the project data, thecustomer152 will often make apayment1916. The third-partynetwork service provider1902 maintains auser database1904 that identifies itsuser base1920 includingusers1906A,1906B . . .1906C.
• Theservice provider1902 may be, for example, an Internet business that provides any of a variety of services to users, such as Internet access, e-mail, web page hosting, domain name hosting, file sharing services or any other Internet-based service. In addition, such Internet-based services may be offered for free or low cost to users, in which case the users have historically agreed to view banner or other advertisements in return for the free or low cost service. However, as stated above, advertising revenue has been subject to diminished pricing and has become an unreliable source of revenue for many Internet-based companies. To facilitate the number of projects that the distributedprocessing system100 can take on and the speed at which these projects can be processed and completed, it is desirable to increase the amount and capabilities of the computing resources available to the distributedprocessing system100. To the extent that the users of theservice provider1902 represent a pool of underutilized resources, these users represent a potentially valuable resource to the distributedprocessing system100.
• According to one or more embodiments, theservice provider1902 may realize value from its user base and thereby monetize this user base by facilitating the use by the distributedprocessing system100 of computing resources related to these users. Thus, for example, in return for free services, the users may agree to have their respective computing resources utilized by the distributedprocessing system100. Theservice provider1902 may then provide to the distributedprocessing system100 the user identifications (IDs)1908 related to its user base in return forrevenue sharing1910. This monetizing architecture thereby enables service providers or other entities that control or have user bases with useful processing capabilities, such as Internet-based service providers, to generate revenue from its user base, particularly in the face of falling revenue from other sources, such as advertising revenue.
• Therevenue sharing1910 may be, for example, a share ofpayment1916 relative to the amount of processing toward theproject1912 that was completed through the use of theuser resources1922 made available throughusers1906A,1906B . . .1906C. It is noted that therevenue sharing1920 may take any desired form, including but not limited to (a) upfront payments based upon attributes of the user base, such as size or processing capabilities, (b) payments based upon the number of users that become members of the distributed processing system, (c) payments based upon the types of projects processed by the user base, or (d) any other desired compensation scheme related to the value of the user base being made available by the third party.
• The monetizing method focuses on the capabilities of the Internet, intranet, wireless or otherwise network connected PCs, internet appliances, notebook computers, servers, storage devices, NAS (Network Attached Storage), or any other connected computing device that could provide any of a number of useful capabilities and that is part of a underutilized user base, such as user bases of Internet-based businesses that rely on advertising or any other method of monetizing their user base in exchange for a valuable service (e.g. free internet access, email, etc.). As discussed above, these useful processing capabilities span the entire range of generic computing subsystems including: Central Processing Unit(s) (CPUs), Digital Signal Processor(s) (DSPs), Graphics Processing Engine(s) (GPEs), Hard Drive(s) (HDs), Memory (MEM), Audio Subsystem(s) (ASs), Communications Subsystem(s) (CSs), Removable Media Types (RMs), or other Add-In/On Accessories (A/OAs) with potentially useful unused capabilities. Market creation and potential compensation for all unused capabilities can be accomplished through the massively parallel distributed software architecture of the distributedprocessing system100. For example, credits (revenues) would be generated each time a unit of work is accomplished by one or more of the subsystems on a user's computing device via a client agent installed on the device to manage, process and complete units of work. The total credits/revenues generated may be, for example, dynamic depending on how many are received. Through this architecture significant revenues may be generated from the user base of the service provider where the service provider may have previously been unable to effectively monetize his user base.
• It is further noted, thatentity1902 may be any entity that has direct or indirect control over a group of users, such that the user's resources may be offered to and utilized by the distributedprocessing system1902. For example, a company with a large group of internal users that are linked to the distributedprocessing system100 through an intranet network of computer systems or computing devices. The user's computing resources may be monetized according to at least one embodiment.
• FIG. 20 is a block diagram representing the components ofclient agent270 which indicates the various responsibilities for those components.Client agent270 includes acore agent component1502, aproject component1504 and auser interface component1510. As discussed above, thecore agent component1502 can provide the base distributed processing functionality for theclient agent270. Theproject component1504 can provide the project-specific functionality for theclient agent270, and theuser interface1510 can provide any desired user viewable information or interaction functionality. These three general components may be modular software components such that theproject component1504, thecore agent component1502, and theuser interface component1510 may be separate software modules that are linked together through appropriate APIs (Application Programming Interface). Thus, each of these components may be designed and developed independently or jointly, as desired. In effect, thecore agent component1502 can provide a backbone upon which is attached theproject component1504, theuser interface component1510, and any other desired components. Thus, when a new project or interface is desired for any givenclient agent270, for example, this component may be efficiently replaced with the new component in a modular fashion still utilizing thecore agent component1502 as the backbone. In addition, each component may be updated and modified without requiring modification or updates to the other component's software code.
• FIG. 20 also depictscustomer152, distributed processing system (DPS)100, andservice provider1902, which communicate with each other through interactions orinterfaces2012 and2014. In this embodiment, thecustomer152 may provide the software development andcode2002 for theproject component1504. The distributedprocessing system100 may provide thecore agent code2008 for thecore agent component1502. And the service provider may provide at least a portion of the interface development andcode2010 for theuser interface component1510. In operation, theworkloads2004 and theresults2006 are typically under the control of the distributedprocessing system100.
• It is noted that this modular architecture facilitates the development of project and interface software code by entities other than the owner of the distributedprocessing system100. For example, with respect toFIG. 19, an Internet-based service provider may have designed and implemented a use interface for its user base, such as a web browser user interface for users of free Internet access services provided by such a service provider. Once thecore agent component1502 is installed on a user's computer, the existing third-party user interface may hook into thecore agent component1502, thereby making the user's resources available to the distributedprocessing system100, while maintaining the user interface the user has come to expect from the service provider. Thus, theservice provider1902 may provide the user interface it desires for the service it is providing, while at the same time monetizing its user base by facilitating its users becoming part of the available resources for the distributedprocessing system100.
CONCLUSION
• Various embodiments enable a server system to consider a variety of priorities in determining how to organize, allocate, and manage client workloads in a distributed processing system.
• Although the subject matter has been described in language specific to structural features and/or methodological steps, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or steps described. Rather, the specific features and steps are disclosed as example forms of implementing the claimed subject matter.

Claims (20)

1. A method, comprising:

generating a capability vector for one or more distributed devices, the capability vector defining a capability of the one or more distributed devices to process workloads;

determining a client priority for one or more client systems, the client priority defining an order that a client system is granted access to submit one or more workloads to a distributed processing system for processing;

assigning a workload priority to one or more workloads, the workload priority defining an order in which a workload is submitted to the distributed processing system for processing; and

distributing one or more workloads from a client system to one or more distributed devices in response to the capability vector, the client priority, and the workload priority.

2. A method as recited inclaim 1, wherein the capability vector is associated with a distributed device's operating performance, data present on a distributed device for processing a workload, and/or application software present on a distributed device for processing a workload.

3. A method as recited inclaim 1, wherein the client priority is based in part on a payment paid by the one or more client systems to access the distributed processing system.

4. A method as recited inclaim 1, wherein the client priority is based in part on an association of one or more client systems with a group of distributed devices.

5. A method as recited inclaim 1, further comprising:

storing the capability vector, the client priority, and the workload priority in a capability database; and

modifying the capability database in response to receiving at least one status input from one or more distributed devices and/or one or more client systems.

6. A method as recited inclaim 5, wherein a status input is received responsive to a change in computing resources associated with the one or more distributed devices.

7. A method as recited inclaim 5, wherein a status input is received responsive to a change in workload priority associated with the one or more client systems.

8. A method as recited inclaim 1, further comprising grouping selected distributed devices from the one or more distributed devices to form one or more host groups, wherein a host group is formed based in part on one or more common capabilities.

9. A method as recited inclaim 8, wherein an incentive is provided to a distributed device to join a host group.

10. A method as recited inclaim 8, further comprising directing a distributed device to allocate a percentage of the distributed device's processing capability between a plurality of project types.

11. A method as recited inclaim 8, further comprising assigning a group value to the capability vector of each distributed device assigned to a host group.

12. A method as recited inclaim 11, wherein the group value is associated with shared resources.

13. Computer readable storage media comprising computer executable instructions comprising:

instructions to generate a capability vector for one or more distributed devices, the capability vector defining at least one capability of the one or more distributed devices to process workloads;

instructions to determine a client priority for one or more client systems, the client priority defining an order that the one or more client systems is granted access to submit one or more workloads to the distributed processing system for processing;

instructions to assign a workload priority to each workload, the workload priority defining an order in which the workload is submitted to the distributed processing system for processing;

instructions to distribute a workload to a distributed device in response to the capability vector, the client priority, and the workload's priority;

instructions to store the capability vector, the client priority, and the workload priority in a capability database; and

instructions to modify the capability database in response to receiving at least one status input from the one or more distributed devices and/or the one or more client systems.

14. Computer readable storage media as recited inclaim 10, wherein a status input is received responsive to a change in computing resources associated with the one or more distributed devices.

15. Computer readable storage media as recited inclaim 10, wherein a status input is received responsive to a change in resource availability associated with the one or more distributed devices.

16. Computer readable storage media as recited inclaim 10, wherein a status input is received responsive to a change in workload priority associated with the one or more client systems.

17. Computer readable storage media to store computer executable instructions comprising:

instructions to monitor a distributed device's resource utilization;

instructions to generate a status for a distributed device, the status describing types of resources, capacities of resources, and utilization of resources;

instructions to modify one or more statuses when resources are added, deleted, and/or availability of a resource changes; and

instructions to send one or more modified statuses to a server system.

18. Computer readable storage media as recited inclaim 17, wherein the one or more modified statuses is sent to the server system responsive to a status request from the server system and/or a status change associated with the distributed device.

19. Computer readable storage media as recited inclaim 18, wherein a status change occurs responsive to a change in computing resources associated with the distributed device.

20. Computer readable storage media as recited inclaim 18, wherein a status change occurs responsive to a change in computing resource availability associated with the distributed device.

Images