Patent Grant
8572236
This invention relates to distributing services in graph-based computations.
Complex business systems typically process data in multiple stages, with the results produced by one stage being fed into the next stage. The overall flow of information through such systems may be described in terms of a directed data flow graph, with vertices in the graph representing components (either data files or processes), and the links or “edges” in the graph indicating flows of data between components.
Graphs also can be used to invoke computations directly. The “CO>OPERATING SYSTEM®” with Graphical Development Environment (GDE) from Ab Initio Software Corporation, Lexington, Mass. embodies such a system. Graphs made in accordance with this system provide methods for getting information into and out of individual processes represented by graph components, for moving information between the processes, and for defining a running order for the processes. This system includes algorithms that choose interprocess communication methods and algorithms that schedule process execution, and also provides for monitoring of the execution of the graph.
A distributed services architecture allows computer programs to access capabilities of other programs through a network interface, such as the world wide web, without having to incorporate the functionalities of those programs into their own operating code.
In general, in one aspect, a service request is processed according to a computation graph associated with the service by receiving inputs for the computation graph from a service client, providing the inputs to the computation graph as records of a data flow, receiving output from the computation graph, and providing the output to the service client.
Implementations may include one or more of the following features.
Receiving the inputs includes identifying the inputs within a service request from the service client, and providing the inputs to the computation graph includes associating a request identifier with the service request, providing a work element including the inputs to the computation graph, associating the request identifier with the work element, and communicating the association of the request identifier with the work element to an output node. Identifying the inputs includes extracting the inputs from an encapsulated request. The encapsulated request includes a SOAP request. Receiving the output includes receiving a work element including outputs from the computation graph, and identifying a request identifier associated with the work element, and providing the output to the service client includes generating a service response corresponding to the service request identified by the request identifier, the service response including the outputs included in the work element. Generating the service response includes embedding the outputs in an encapsulated response. The encapsulated response includes a SOAP response.
A second service request is processed according to the computation graph by receiving second inputs for the computation graph from a second service client, providing the second inputs to the computation graph as records of a second data flow, receiving second output from the computation graph, and providing the second output to the second service client; and in the computation graph, processing the first inputs and the second inputs concurrently. Processing the first inputs and the second inputs concurrently includes identifying a first subgraph applicable to the first inputs, processing the first inputs in the first subgraph, identifying a second subgraph applicable to the second inputs, processing the second inputs in the second subgraph, receiving first outputs from the first subgraph, and receiving second outputs from the second subgraph, in which the first subgraph and the second subgraph are executed simultaneously. Processing the first inputs and the second inputs concurrently includes identifying a first instance of a subgraph as available, processing the first inputs in the first instance of the subgraph, identifying a second instance of the subgraph as available, processing the second inputs in the second instance of the subgraph, receiving first outputs from the first instance, and receiving second outputs from the second instance, in which the first instance of the subgraph and the second instance of the subgraph are executed simultaneously.
In general, in one aspect, a data flow is processed in a graph-based computation by identifying a first graph applicable to a first record in the data flow, identifying a first subgraph applicable to a first work element in the first record, processing the first work element in the first subgraph, identifying a second subgraph applicable to a second work element in the first record, processing the second work element in the second subgraph, receiving a first output work element from the first subgraph, receiving a second output work element from the second subgraph, associating the first output work element with the first record, and associating the second output work element with the first record, in which the first subgraph and the second subgraph are executed at least partially simultaneously.
In general, in one aspect, a data flow is processed in a graph-based computation by identifying a first graph applicable to a first record in the data flow, identifying a first instance of a subgraph as available, processing a first work element in the first record in the first instance of the subgraph, identifying a second instance of the subgraph as available, processing a second work element in the first record in the second instance of the subgraph, receiving a first output work element from the first instance, receiving a second output work element from the second instance, associating the first output work element with the first record, and associating the second output work element with the first record, in which the first instance of the subgraph and the second instance of the subgraph are executed at least partially simultaneously.
Other features and advantages of the invention will be apparent from the description and the claims.
b show graphs for providing distributed services.
In some examples, there are two ways that graph-based computations interact with distributed services. In some examples, a graph calls a service and relates to the service as if the service were a node in the graph. In other examples, an application relates to a graph such that the graph provides a service. The two cases may be carried out simultaneously. For example, a graph that is providing services to one application may itself call another service to perform part of its computation. The service that one graph calls may be provided by another graph, such that the first graph calls the second rather than containing it as a subgraph, for example.
In some examples, as shown in
The service node 102 receives inputs 110 from other graph nodes, e.g., a node 112, in the native format of the graph-based computation 100, such as a record described in a data description language like Ab Initio's DML. The node 102 then formats its inputs 110 into a format appropriate for the type of web service it is using, such as a SOAP- or other type of encapsulation-based request, and transmits a service request 114 to the application server 106. A URL in the request 114 indicates which application server 106, plug in 104, and service 108 to send the request to. The request 114 can be sent in various ways, including by HTTP and RPC (remote procedure call). The plug-in 104 on the server 106 reformats the incoming request 114 as appropriate and transmits an appropriate call 116 to the external service 108.
After the service 108 performs whatever operations it was called for, it returns output 118 to the plug-in 104. The plug-in 104 reformats the output 118 of the service 108 into an appropriate service response 120 to return to the service node 102. The service node 102 reformats the response 120 into an appropriate output format and passes output 121 on to downstream graph nodes, e.g., a node 122. From the point of view of the other graph nodes 112, 122, 124, 126, the distributed service node 102 can participate in the graph computation 100 just as any other node. The structure and organization of the server 106, such as the nature of the plug in 104 and how it communicates to the service 108 (or whether, for example, a plug-in 104 is used at all) do not matter to the graph nodes, as long as the service node 102 receives the type of response 120 it is expecting. The server 106, plug-in 104, and service 108 may themselves be a graph-based computation server, plug-in, and graph, respectively, as described below.
In some examples, the graph 100 is pipelined, meaning that each node operates on a separate record, or set of records, of a larger flow of input records in sequential phases of the pipeline. In a given phase, each subsequent node works on a record, or set of records, that it received from a previous node, while each previous node works on the next record or set of records in the input flow. The service node 102 modifies this by accepting additional records and dispatching the work required by each of them to the service 108 before responses corresponding to previous service requests have been received. This is possible where, for example, the service 108 is capable of processing multiple concurrent requests. Records may be composed of separate or separable work elements, which may be handled in pipelined stages as described for records.
If the different requests 114 take different amounts of time to process, the service node 102 may receive the responses in a different order than its inputs 110 were received. This allows multiple services having different processing times to be accessed by a single node 102. A service node such as 102 may optionally be configured to provide its outputs 121 in a different order than its inputs 110 were received. To allow such concurrent processing without using a multi-threaded process for the service node 102 and without including data describing the entire state of the graph 100 in each request 114 (so that it can be used upon its return), the service node 102 keeps track of which requests 114 are outstanding and matches them to incoming responses 120. In some examples, the graph is pipelined such that each record or work element is handled by one node at a time and separate nodes handle separate records or work elements, but multiple records or work elements may be allocated to the single service node 102 at any given time.
In a pipelined graph, the service node 102 treats each record, e.g., each record of a data flow, as an individual transaction. The corresponding request is generated based on the fields in that single input record. Some of the fields may define attributes of the request, such as the URL, while others are passed on as embedded data.
In some examples, as shown in a process 150 in
In some examples, a graph-based computation is provided as a network service. We refer to the graph that provides a network service as a provider graph. As shown in
The web services plug-in 210 is an application that processes requests and responses between web services clients and the graphs providing the services. The plug-in 210 is installed and executed on the application server 208. The web services plug-in may communicate with a set of several web services provider graphs. Each of the graphs is included in a services list accessed by the plug-in. The plug-in uses a URL, for example, in a client HTTP request message, to direct the request to a particular graph.
The plug-in 210 formats the service request 204 into an appropriate request 212 for graph processing, such as an RPC request, and forwards this request 212 to a graph server 214 which is hosting the graph 206. In some examples, RPC calls are made straight through from the client 200 to the provider graph 206. At the graph server 214, the inputs 216 included in the request 212 are provided to the graph 206, which is executed to produce outputs 218. The graph server 214 sends the outputs 218 in an appropriate form of graph output 220, such as an RPC response, to the application server 208 and its web service plug-in 210. The plug-in 210 reformats the graph output 220 into a web-services response 222 in a complementary format to the original request 204. The outputs are returned to the client or other entity that called the service in the usual manner of the appropriate network service. The process on the client system 202 then handles the response as it would any other web-based or network service response.
The graph may process inputs using any or all of pipeline parallelism, component parallelism, and data parallelism. That is, multiple sets of inputs may be received, and these may be interleaved in pipeline stages, or they may be processed concurrently by different groups of components, perhaps after partitioning.
In some examples, the public or external interface to a provider graph is defined by a WSDL (Web Services Description Language) file. A WSDL file contains XML-based descriptions of all the messages and operations needed to interact with the provider graph. WSDL files may be defined by organizations, industries, or any other body, for accessing web services relevant to that body. WSDL files may also be customized for a particular implementation. In some examples, a WSDL file is generated from record formats and type definitions of components in subgraphs related to the provider graph. In some examples, a single provider graph includes multiple subgraphs that perform separate services. The WSDL file crated from such subgraphs allows a client to access each of the services provided by the provider graph's subgraphs.
Within the provider graph, the incoming service request input is converted into the internal language of the service-providing subgraph. For example, as shown in
Once the message 306 is translated, a Partition by Action Type node 302 interprets the action specified in the message 306 and sends the record 308 in the data payload to the appropriate action processor 300a or 300b. In some examples, the partition node 302 does not accept the next incoming message 307 until a gather node 304 informs it that the work for the previous record 308 has been output from the action processor 300a or 300b.
In some examples, to achieve concurrency in handling of messages, e.g., service requests, the partition node 302 is configured to partition the action records corresponding to the service requests. This allows multiple requests to be handled simultaneously, if they don't require the same action processor. For example, if a first record 308 requires the action processor 300a, and a second record 310 requires the action processor 300b, the partition node 302 sends the second record 310 to the action processor 300b without waiting for the gather node 304 to report that the first record 308 has completed its processing.
The gather node 304 receives the processed records 308 and 310 from the action processors 300a and 300b and packages them into response messages 312 and 313 containing the processed records 308′ or 310′ as data payloads. If the partition node 302 is partitioning transactions, so that multiple records may be processed at once, it is possible that the processed records 308′ or 310′ may be received by the gather node 304 in a different order than their corresponding request messages 306 and 307 were received. An interface 315 between the partition node 302 and gather node 304 allows the gather node to associate the output records 308′, 310′ with the corresponding inputs 308, 310 so that the reply messages 312 and 313 are sent back to the correct client.
The action processors 300a, 300b can take any form. As noted above, the records 308, 310 may be in any format, e.g., DML-described. In such a case, the interface layer 301 translates the incoming language, such as SOAP, into the DML format used by the action processors. In some examples, an intermediate language may be used to express the inputs between the interface layer 301 and the partition and gather nodes 302, 304.
In some examples, when gather node 304 is configured to receive records 308′ and 310′ in the order they arrive, processing of actions may be further partitioned within the action processors, as shown for the action processor 300b in
The output records 308′ and 310′ and output messages 312 and 313 may be returned in a different order than the input requests 306 and 307 were received. For example, if the input record 310 contained fewer work elements (described below) than the record 308, then the output response 313 may be returned first even though it corresponds to the later-received request 307. The output order may also vary based on the complexity of the input data, if the action processors 300a, 300b work in such a way that records may be processed out of order or at different rates depending on their complexity, or other factors
In some examples, as shown in
In some examples, once actions are partitioned and unwrapped into work elements, the work elements are also processed concurrently, as shown in
As the work elements 408a′-408e′ leave the processors 500a, 500b, they are accumulated by the gather node 504. Whenever all the work elements from one record 406 have been collected, they are combined back into the output record 412. Communication 515 between the partition node 502 and the gather node 504 allows the partition node 502 to track which work element processors are available for additional work elements.
Techniques for handling multiple concurrent data flows are also described in U.S. Pat. No. 6,584,581, issued Jun. 24, 2003, and U.S. Pat. No. 6,654,907, issued Nov. 25, 2003, the entire contents of which are hereby incorporated by reference.
b shows details of how the work element processor 500b operates and how the subgraph of
In previous systems, a partition component similar to node 552 would have partitioned a single work element 514 into separate components and sent one component to each instance 550i (where i is a, b, or c). The gather node 554 would have waited until each instance returned an output before combining the outputs into a single output for that work element, eliminating the possibility of concurrency. To maintain the pipeline, the partition node may have sent a delimiter through each instance after the work element so that the gather component would know that the work element was completed when the delimiter was received from each instance. In some examples, the delimiters would be sent after a set of multiple work elements that are handled by a given node in a given phase of the pipeline.
The partition by pool node 552, in contrast, sends the entire work element 408a to a single instance 550a for processing. The gather node 554 is configured to expect the output 408a′ for that work element from only that one instance 550a. When the expected output 408a′ is received (in some examples, followed by a delimiter), the gather node 544 provides the output to the next node 504. The gather node 554 does not wait for work element components or delimiters from the other instances 550b and 550c. This allows work element processor 500b to only use the instances 550i that are really needed for a given work element. As with the partitioning of transactional records by pool in
Partitioning work elements between pools also allows the outputs 408a′, 408b′, 408c′ to be returned out of order, relative to the inputs 408a, 408b, 408c. For example, if the data in the work element 408a causes the instance 550a to take longer process that work element than the instance 550b takes to process the work element 408c, the gather node 554 will receive the output 408c′ before it receives the output 408a′. Because the two work elements 408a and 408c are independent of each other, the gather node 554 can send the output 408c′ to the next node 504 without waiting for the output 408a′. This assumes that the other stages (e.g., partion/gather pair 502/504 and unwrap/wrap pair 402/404) are configured to accommodate changes in the order of work elements within the work element processor 500b. In some examples, it may be necessary or desirable to maintain the order, so the gather node 554 will hold the output 408c′ until the output 408a′ is received, and release them to the next node 504 in order. There are still advantages in such a computation, as the instances 550a, 550b, 550c are able to process several work elements simultaneously. This may be useful, for example, if the processing time of the subgraph 550 is longer than that of other nodes or subgraphs.
Partitioning transactions by action type and by pool and partitioning work elements by type and by pool all allow the graph to operate with internal parallelism—that is, with concurrent processing of related or unrelated work elements by different processing elements.
Partitioning by pool, partitioning by type, and other methods of processing requests and their work elements concurrently can be implemented in provider graphs so that the services the graphs provide can themselves handle concurrent inputs. For example, the graph 501 in
In some examples, the subscribe and publish nodes 570 and 572 assume that individual requests 574 are independent and the order in which responses 576 are returned does not matter, so the partition and gather nodes 402, 404 in the graph 501 are configured to not enforce order. If order does matter, either or both of the subscribe node 570 and the publish node 572 may make sure that responses 576 are returned the same order as the corresponding inputs 574 were received.
The techniques described herein may be implemented in one or more computer programs executing on one or more programmable computer systems each comprising at least one processor, at least one data storage system (including volatile and nonvolatile memory and/or storage elements), at least one input device or port, and at least one output device or port. The program code is executed on the processors to perform the functions described herein.
Each such program may be implemented in any desired computer language (including machine, assembly, or high level procedural, logical, or object oriented programming languages) to communicate with a computer system. In any case, the language may be a compiled or interpreted language.
Each such computer program is preferably stored on a storage media or device (e.g., ROM, CD-ROM, or magnetic or optical media) readable by a general or special purpose programmable computer system, for configuring and operating the computer when the storage media or device is read by the computer system to perform the procedures described herein. The inventive system may also be considered to be implemented as a computer-readable storage medium, configured with a computer program, where the storage medium so configured causes a computer system to operate in a specific and predefined manner to perform the functions described herein.
A number of embodiments of the present invention have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. For example, a number of the function steps described above may be performed in a different order without substantially affecting overall processing.
Other implementations are within the scope of the following claims and other claims to which the applicant may be entitled.
This application claims priority to Provisional Patent Application Ser. No. 60/836,745, filed on Aug. 10, 2006, the entire contents of which are hereby incorporated by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 3662343 | Goldstein et al. | May 1972 | A |
| 3662401 | Collins et al. | May 1972 | A |
| 4922418 | Dolecek | May 1990 | A |
| 4972314 | Getzinger et al. | Nov 1990 | A |
| 5127104 | Dennis | Jun 1992 | A |
| 5276899 | Neches | Jan 1994 | A |
| 5280619 | Wang | Jan 1994 | A |
| 5301336 | Kodosky | Apr 1994 | A |
| 5323452 | Dickman et al. | Jun 1994 | A |
| 5333319 | Silen | Jul 1994 | A |
| 5357632 | Pian et al. | Oct 1994 | A |
| 5495590 | Comfort et al. | Feb 1996 | A |
| 5630047 | Wang | May 1997 | A |
| 5692168 | McMahan | Nov 1997 | A |
| 5701400 | Amardo | Dec 1997 | A |
| 5712971 | Stanfill et al. | Jan 1998 | A |
| 5745778 | Alfieri | Apr 1998 | A |
| 5802267 | Shirakihara et al. | Sep 1998 | A |
| 5805462 | Poirot et al. | Sep 1998 | A |
| 5857204 | Lordi et al. | Jan 1999 | A |
| 5923832 | Shirakihara et al. | Jul 1999 | A |
| 5924095 | White | Jul 1999 | A |
| 5930794 | Linenbach et al. | Jul 1999 | A |
| 5933640 | Dion | Aug 1999 | A |
| 5966072 | Stanfill et al. | Oct 1999 | A |
| 5999729 | Tabloski, Jr. et al. | Dec 1999 | A |
| 6006242 | Poole et al. | Dec 1999 | A |
| 6012094 | Leymann | Jan 2000 | A |
| 6014670 | Zamanian et al. | Jan 2000 | A |
| 6016516 | Horikiri | Jan 2000 | A |
| 6032158 | Mukhopadhhyay et al. | Feb 2000 | A |
| 6038558 | Powers et al. | Mar 2000 | A |
| 6044211 | Jain | Mar 2000 | A |
| 6044374 | Nesamoney et al. | Mar 2000 | A |
| 6044394 | Cadden et al. | Mar 2000 | A |
| 6088716 | Stanfill et al. | Jul 2000 | A |
| 6145017 | Ghaffari | Nov 2000 | A |
| 6173276 | Kant et al. | Jan 2001 | B1 |
| 6208345 | Sheard et al. | Mar 2001 | B1 |
| 6256637 | Venkatesh et al. | Jul 2001 | B1 |
| 6259988 | Galkowski et al. | Jul 2001 | B1 |
| 6272650 | Meyer et al. | Aug 2001 | B1 |
| 6301601 | Helland | Oct 2001 | B1 |
| 6314114 | Coyle et al. | Nov 2001 | B1 |
| 6324437 | Frankel et al. | Nov 2001 | B1 |
| 6330008 | Razdow et al. | Dec 2001 | B1 |
| 6339775 | Zamanian et al. | Jan 2002 | B1 |
| 6400996 | Hoffberg et al. | Jun 2002 | B1 |
| 6401216 | Meth et al. | Jun 2002 | B1 |
| 6437796 | Sowizral et al. | Aug 2002 | B2 |
| 6449711 | Week | Sep 2002 | B1 |
| 6480876 | Rehg et al. | Nov 2002 | B2 |
| 6496961 | Gupta et al. | Dec 2002 | B2 |
| 6538651 | Hayman et al. | Mar 2003 | B1 |
| 6584581 | Bay et al. | Jun 2003 | B1 |
| 6608628 | Ross et al. | Aug 2003 | B1 |
| 6611862 | Reisman | Aug 2003 | B2 |
| 6651234 | Gupta et al. | Nov 2003 | B2 |
| 6654907 | Stanfill et al. | Nov 2003 | B2 |
| 6658464 | Reisman | Dec 2003 | B2 |
| 6715145 | Bowman-Amuah | Mar 2004 | B1 |
| 6728879 | Atkinson | Apr 2004 | B1 |
| 6813761 | Das et al. | Nov 2004 | B1 |
| 6816825 | Ashar et al. | Nov 2004 | B1 |
| 6832369 | Kryka et al. | Dec 2004 | B1 |
| 6848100 | Wu et al. | Jan 2005 | B1 |
| 6879946 | Rong et al. | Apr 2005 | B2 |
| 7062483 | Ferrari et al. | Jun 2006 | B2 |
| 7082604 | Schneiderman | Jul 2006 | B2 |
| 7085426 | August | Aug 2006 | B2 |
| 7103597 | McGovern | Sep 2006 | B2 |
| 7103620 | Kunz et al. | Sep 2006 | B2 |
| 7130484 | August | Oct 2006 | B2 |
| 7137116 | Parkes et al. | Nov 2006 | B2 |
| 7164422 | Wholey et al. | Jan 2007 | B1 |
| 7165030 | Yi et al. | Jan 2007 | B2 |
| 7167850 | Stanfill | Jan 2007 | B2 |
| 7316001 | Gold et al. | Jan 2008 | B2 |
| 7356819 | Ricart et al. | Apr 2008 | B1 |
| 7398514 | Ulrich et al. | Jul 2008 | B2 |
| 7417645 | Beda et al. | Aug 2008 | B2 |
| 7457984 | Kutan | Nov 2008 | B2 |
| 7467383 | Inchingolo et al. | Dec 2008 | B2 |
| 7505975 | Luo | Mar 2009 | B2 |
| 7577628 | Stanfill | Aug 2009 | B2 |
| 7636699 | Stanfill | Dec 2009 | B2 |
| 7716630 | Wholey et al. | May 2010 | B2 |
| 7840949 | Schumacher et al. | Nov 2010 | B2 |
| 7870556 | Wholey et al. | Jan 2011 | B2 |
| 7877350 | Stanfill et al. | Jan 2011 | B2 |
| 7979479 | Staebler et al. | Jul 2011 | B2 |
| 20010055019 | Sowizral et al. | Dec 2001 | A1 |
| 20020080181 | Razdow et al. | Jun 2002 | A1 |
| 20020087921 | Rodriguez | Jul 2002 | A1 |
| 20020091747 | Rehg et al. | Jul 2002 | A1 |
| 20020091748 | Rehg et al. | Jul 2002 | A1 |
| 20020111876 | Rudraraju et al. | Aug 2002 | A1 |
| 20020129340 | Tuttle | Sep 2002 | A1 |
| 20020147745 | Houben et al. | Oct 2002 | A1 |
| 20020184616 | Chessell et al. | Dec 2002 | A1 |
| 20030004771 | Yaung | Jan 2003 | A1 |
| 20030023413 | Srinivasa | Jan 2003 | A1 |
| 20030033432 | Simpson et al. | Feb 2003 | A1 |
| 20030091055 | Craddock et al. | May 2003 | A1 |
| 20030126240 | Vosseler | Jul 2003 | A1 |
| 20030204804 | Petri et al. | Oct 2003 | A1 |
| 20040006745 | Van Helden et al. | Jan 2004 | A1 |
| 20040041838 | Adusumilli et al. | Mar 2004 | A1 |
| 20040073529 | Stanfill | Apr 2004 | A1 |
| 20040093559 | Amaru et al. | May 2004 | A1 |
| 20040098452 | Brown et al. | May 2004 | A1 |
| 20040107414 | Bronicki et al. | Jun 2004 | A1 |
| 20040111469 | Manion et al. | Jun 2004 | A1 |
| 20040148373 | Childress et al. | Jul 2004 | A1 |
| 20040177099 | Ganesh et al. | Sep 2004 | A1 |
| 20040205726 | Chedgey et al. | Oct 2004 | A1 |
| 20040207665 | Mathur | Oct 2004 | A1 |
| 20040210831 | Feng et al. | Oct 2004 | A1 |
| 20040225657 | Sarkar | Nov 2004 | A1 |
| 20040260590 | Golani et al. | Dec 2004 | A1 |
| 20050021689 | Marvin et al. | Jan 2005 | A1 |
| 20050034112 | Stanfill | Feb 2005 | A1 |
| 20050039176 | Fournie | Feb 2005 | A1 |
| 20050059046 | Labrenz et al. | Mar 2005 | A1 |
| 20050086360 | Mamou et al. | Apr 2005 | A1 |
| 20050097561 | Schumacher et al. | May 2005 | A1 |
| 20050102670 | Bretl et al. | May 2005 | A1 |
| 20050144277 | Flurry et al. | Jun 2005 | A1 |
| 20050144596 | McCullough et al. | Jun 2005 | A1 |
| 20050149935 | Benedetti | Jul 2005 | A1 |
| 20050177531 | Bracewell | Aug 2005 | A1 |
| 20050193056 | Schaefer et al. | Sep 2005 | A1 |
| 20050216421 | Barry et al. | Sep 2005 | A1 |
| 20050240621 | Robertson et al. | Oct 2005 | A1 |
| 20050262470 | Gavrilov | Nov 2005 | A1 |
| 20050289527 | Illowsky et al. | Dec 2005 | A1 |
| 20060095722 | Biles et al. | May 2006 | A1 |
| 20060098017 | Tarditi et al. | May 2006 | A1 |
| 20060206872 | Krishnaswamy | Sep 2006 | A1 |
| 20060294150 | Stanfill et al. | Dec 2006 | A1 |
| 20060294459 | Davis et al. | Dec 2006 | A1 |
| 20070011668 | Wholey et al. | Jan 2007 | A1 |
| 20070022077 | Stanfill | Jan 2007 | A1 |
| 20070035543 | David et al. | Feb 2007 | A1 |
| 20070094211 | Sun et al. | Apr 2007 | A1 |
| 20070118839 | Berstis et al. | May 2007 | A1 |
| 20070139441 | Lucas et al. | Jun 2007 | A1 |
| 20070143360 | Harris et al. | Jun 2007 | A1 |
| 20070150429 | Huelsman et al. | Jun 2007 | A1 |
| 20070174185 | McGovern | Jul 2007 | A1 |
| 20070239766 | Papaefstathiou et al. | Oct 2007 | A1 |
| 20070285440 | MacInnis et al. | Dec 2007 | A1 |
| 20080126755 | Wu et al. | May 2008 | A1 |
| 20080288608 | Johnson | Nov 2008 | A1 |
| 20080294615 | Furuya et al. | Nov 2008 | A1 |
| 20090030863 | Stanfill et al. | Jan 2009 | A1 |
| 20090182728 | Anderson | Jul 2009 | A1 |
| 20090193417 | Kahlon | Jul 2009 | A1 |
| 20090224941 | Kansal et al. | Sep 2009 | A1 |
| 20090327196 | Studer et al. | Dec 2009 | A1 |
| 20100070955 | Kahlon | Mar 2010 | A1 |
| 20100169137 | Jastrebski et al. | Jul 2010 | A1 |
| 20100174694 | Staebler et al. | Jul 2010 | A1 |
| 20100180344 | Malyshev et al. | Jul 2010 | A1 |
| 20100211953 | Wakeling et al. | Aug 2010 | A1 |
| 20100218031 | Agarwal et al. | Aug 2010 | A1 |
| 20100281488 | Krishnamurthy et al. | Nov 2010 | A1 |
| 20110078500 | Douros et al. | Mar 2011 | A1 |
| 20110093433 | Stanfill et al. | Apr 2011 | A1 |
| 20120054255 | Buxbaum | Mar 2012 | A1 |
| Number | Date | Country |
|---|---|---|
| 64-013189 | Jan 1989 | JP |
| 06-236276 | Aug 1994 | JP |
| 08-278892 | Oct 1996 | JP |
| 08-305576 | Nov 1996 | JP |
| 63-231613 | Sep 1998 | JP |
| 11-184766 | Jul 1999 | JP |
| 2000-99317 | Apr 2000 | JP |
| 2006-504160 | Feb 2006 | JP |
| WO 9800791 | Jan 1998 | WO |
| WO 0211344 | Feb 2002 | WO |
| WO 2005001687 | Jan 2005 | WO |
| WO 2005086906 | Sep 2005 | WO |
| WO 2008124319 | Oct 2008 | WO |
| WO 2009039352 | Mar 2009 | WO |
| Entry |
|---|
| International Search Report dated Sep. 16, 2008 in PCT Application No. PCT/US07/75576, 2 pages. |
| Supplementary European Search Report issued in EP application No. 07813940, Nov. 26, 2009, 7 pages. |
| Krsul, Ivan et al., “VMPlants: Providing and Managing Virtual Machine Execution Environments for Grid Computing.” Proceedings of the ACM/IEEE SC2004 Conference on Supercomputing, 2001, Nov. 6-12, 2004, 12 pages. |
| Vajracharya, Suvas et al., “Asynchronous Resource Management.” Proceedings of the 15th International Parallel and Distributed Processing Symposium, Apr. 2001, 10 pages. |
| Control-M; New Dimension Software. User Manual. New Dimension Software Ltd., 1999. |
| Romberg, M., “UNICORE: Beyond Web-based Job-Submission,” Proceedings of the 42nd Cray User Group Conference, Noordwijk (May 22-26, 2000). |
| “Unicenter AutoSys Job Management,” Computer Associates, Copyright 2001. |
| Babaoglu, O et al., “Mapping parallel computations onto distributed systems in Paralex” Compuero '91. Advanced Computer Technology, Reliable Systems and Applications. 5th Annual European Computer Conference. Proceedings. Bologna, Italy May 13-16, 1991, Los Alamitos, CA, USA, IEEE Comput. Soc, US, pp. 123-130. |
| Baer, J.L. et al., “Legality and Other Properties of Graph Models of Computations.” Journal of the Association for Computing Machinery, vol. 17, No. 3, Jul. 1970, pp. 543-554. |
| Bookstein, A. et al., “Modeling Word Occurrences for the Compression of Concordances.” ACM Transactions on Information Systems, vol. 15, No. 3, Jul. 1997, pp. 254-290. |
| Burch, J.R. et al., “Sequential circuit verification using symbolic model checking.” In Design Automation Conference, 1990, Proceedings of the 27th ACM/IEEE. Jun. 24-28, 1990, pp. 46-51. |
| Cytron, Ron et al., “Efficiently Computing Static Single Assignment Form and the Control Dependence Graph.” ACM Transactions on Programming Languages and Systems, vol. 13, No. 4, Oct. 1991, pp. 451-490. |
| Ebert, Jurgen et al., “A Declarative Approach to Graph-Based Modeling.” Workshop on Graph-Theoretic Concepts in Computer Science, 1994, pp. 1-19. |
| European Search Report issued in application No. EP10003554, dated Sep. 24, 2010, 7 pages. |
| Gamma et al. “Design Patterns: Elements of Reusable Object-Oriented Software”, Sep. 1999. |
| Guyer et al., “Finding Your Cronies: Static Analysis for Dynamic Object Colocation.” Oct. 2004, ACM, pp. 237-250. |
| Grove et al., “A Framework for Call Graph Construction Algorithms.” Nov. 2001, ACM TOPLAS, vol. 23, Issue 6, pp. 685-746. |
| Herniter, Marc E., “Schematic Capture with MicroSim PSpice,” 2nd Edition, Prentice Hall, Upper Saddle River, N.J., 1996, pp. 51-52, 255-280, 292-297. |
| International Search Report & Written Opinion issued in PCT application No. PCT/US01/23552, mailed Jan. 24, 2002, 5 pages. |
| International Search Report & Written Opinion issued in PCT application No. PCT/US06/24957, dated Jan. 17, 2008, 14 pages. |
| International Search Report & Written Opinion issued in PCT application No. PCT/US08/71206, mailed Oct. 22, 2008, 12 pages. |
| International Search Report & Written Opinion issued in PCT application No. PCT/US10/49966, dated Nov. 23, 2010, 8 pages. |
| International Search Report & Written Opinion received in PCT application No. PCT/US10/24036, mailed Mar. 23, 2010, 11 pages. |
| Jawadi, Ramamohanrao et al., “A Graph-based Transaction Model for Active Databases and its Parallel Implementation.” U. Florida Tech. Rep TR94-0003, 1994, pp. 1-29. |
| Just et al., “Review and Analysis of Synthetic Diversity for Breaking Monocultures.” Oct. 2004, ACM, pp. 23-32. |
| Kebschull, U. et al., “Efficient Graph-Based Computation and Manipulation of Functional Decision Diagrams.” University of Tubingen, 1993 IEEE, pp. 278-282. |
| Krahmer et al., “Graph-Based Generation of Referring Expressions.” Mar. 2003, MIT Press, vol. 29, No. 1, pp. 53-72. |
| Li, Xiqing et al., “A Practical External Sort for Shared Disk MPPs.” Proceedings of Supercomputing '93, 1993, 24 pages. |
| Martin, David et al., “Models of Computations and Systems—Evaluation of Vertex Probabilities in Graph Models of Computations.” Journal of the Association for Computing Machinery, vol. 14, No. 2, Apr. 1967, pp. 281-299. |
| Ou, Chao-Wei et al., “Architecture-Independent Locality-Improving Transformations of Computational Graphs Embedded in k-Dimensions.” Proceedings of the 9th International Conference on Supercomputing, 1995, pp. 289-298. |
| “RASSP Data Flow Graph Design Application Note.” International Conference on Parallel Processing, Dec. 2000, Retrieved from Internet <http://www.atl.external.lmco.com/projects/rassp/RASSP—legacy/appnotes/FLOW/APNOTE—FLOW—02 >, 5 pages. |
| Stanfill, Craig, “Massively Parallel Information Retrieval for Wide Area Information Servers.” 1991 IEEE International Conference on Systems, Man and Cybernetics, Oct. 1991, pp. 679-682. |
| Stanfill, Craig et al., “Parallel Free-Text Search on the Connection Machine System.” Communications of the ACM, vol. 29, No. 12, Dec. 1986, pp. 1229-1239. |
| Stanfill, Craig, “The Marriage of Parallel Computing and Information Retrieval.” IEE Colloquium on Parallel Techniques for Information Retrieval, Apr. 1989, 5 pages. |
| Supplemental European Search Report issued in application No. EP08796632, dated Sep. 24, 2010, 6 pages. |
| Wah, B.W. et al., “Report on Workshop on High Performance Computing and Communications for Grand Challenge Applications: Computer Vision, Speech and Natural Language Processing, and Artificial Intelligence.” IEEE Transactions on Knowledge and Data Engineering, vol. 5, No. 1, Feb. 1993, 138-154. |
| International Search Report & Written Opinion received in PCT application No. PCT/US2011/040440, mailed Oct. 12, 2011, 13 pages. |
| Japanese Office Action, with English Translation, JP application No. 2009-523997, mailed Oct. 23, 2012, 7 pages. |
| Japanese Office Action, with English Translation, JP application No. 2008-519474, mailed Sep. 25, 2012, 8 pages. |
| Shoten, Iwanami, “Encyclopedic Dictionary of Computer Science,” (with English Translation), May 25, 1990, p. 741. |
| Supplemental European Search Report issued in application No. EP06774092, dated Dec. 19, 2012, 5 pages. |
| “Topological sorting,” Wikipedia, accessed Dec. 10, 2012, 2 pages. |
| Japanese Office Action, with English Translation, JP application No. 2011-000948, mailed Jan. 8, 2013, 11 pages. |
| European Search Report issued in application No. EP10741775, dated Nov. 14, 2012, 4 pages. |
| Russell, Nick, et al., “Workflow Control-Flow Patterns A Revised View,” Workflow Patterns Initiative, 2006, pp. 1-134. |
| van der Aalst, W.M.P., et al., “Workflow Patterns,” Distributed and Parallel Databases, 14, 5-51, 2003. |
| Dillon, Laura K., et al., “Inference Graphs: A Computational Structure Supporting Generation of Customizable and Correct Analysis Components,” IEEE Transactions on Software Engineering, vol. 29, No. 2, Feb. 2003, pp. 133-150. |
| Evripidou, Paraskevas, et al., “Incorporating input/output operations into dynamic data-flow graphs,” Parallel Computing 21 (1995) 1285-1311. |
| Extended European Search Report, EP 12165575, mailed May 10, 2013, 9 pages. |
| Frankl, Phyllis G., et al., “An Applicable Family of Data Flow Testing Criteria,” IEEE Transactions on Sofrware Engineering, vol. 14, No. 10, Oct. 1988, pp. 1483-1498. |
| Whiting, Paul G., et al., “A History of Data-Flow Languages,” IEEE Annals of the History of Computing, vol. 16, No. 4, 1994, pp. 38-59. |
| Japanese Office Action, with English Translation, JP application No. 2010-518415, mailed Feb. 21, 2013, 11 pages. |
| “Visual Lint: Squash Bugs Early with Interactive C/C++, C# and Java Code Analysis for Microsoft Visual Studio and Eclipse,” [retrieved from the internet Dec. 3, 2012: www.riverblade.co.uk/products/visual—lint.] (2 pages). |
| Number | Date | Country | |
|---|---|---|---|
| 20080049022 A1 | Feb 2008 | US |
| Number | Date | Country | |
|---|---|---|---|
| 60836745 | Aug 2006 | US |
