Patent application number | Description | Published |
20080263329 | Parallel-Prefix Broadcast for a Parallel-Prefix Operation on a Parallel Computer - A parallel-prefix broadcast for a parallel-prefix operation on a parallel computer includes: configuring, on each node, a parallel-prefix contribution buffer for storing the node's parallel-prefix contribution; configuring, on each node, a parallel-prefix results buffer for storing results of a operation, the results buffer having a position for each node that corresponds to node's rank; and repeatedly for each position in the results buffer: processing in parallel by each node, including: determining, by the node, whether the current position in the results buffer is to include the node's contribution, if the current position is not to include the contribution, contributing the identity element, and if the current position is to include the contribution, contributing the contribution, performing, by each node, the operation using the contributed identity elements and the contributed contributions, yielding a result from the operation, and storing, by each node, the result in the position in the results buffer. | 10-23-2008 |
20080263386 | DYNAMICALLY REROUTING NODE TRAFFIC ON A MASSIVELY PARALLEL COMPUTER SYSTEM USING HINT BITS - A method and apparatus for dynamically rerouting node processes on the compute nodes of a massively parallel computer system using hint bits to route around failed nodes or congested networks without restarting applications executing on the system. When a node has a failure or there are indications that it may fail, the application software on the system is suspended while the data on the failed node is moved to a backup node. The torus network traffic is routed around the failed node and traffic for the failed node is rerouted to the backup node. The application can then resume operation without restarting from the beginning. | 10-23-2008 |
20080263387 | FAULT RECOVERY ON A PARALLEL COMPUTER SYSTEM WITH A TORUS NETWORK - An apparatus and method for overcoming a torus network failure in a parallel computer system. A mesh routing mechanism in the service node of the computer system configures the nodes from a torus to a mesh network when a failure occurs in the torus network. The mesh routing mechanism takes advantage of cutoff registers in each node to route node to node data transfers around the faulty node or network connection. | 10-23-2008 |
20080288746 | Executing Multiple Instructions Multiple Data ('MIMD') Programs on a Single Instruction Multiple Data ('SIMD') Machine - Executing MIMD programs on a SIMD machine, the SIMD machine including a plurality of compute nodes, each compute node capable of executing only a single thread of execution, the compute nodes initially configured exclusively for SIMD operations, the SIMD machine further comprising a data communications network, the network comprising synchronous data communications links among the compute nodes, including establishing one or more SIMD partitions, booting one or more SIMD partitions in MIMD mode; establishing a MIMD partition; executing by launcher programs a plurality of MIMD programs on two or more of the compute nodes of the MIMD partition; and re-executing a launcher program by an operating system on a compute node in the MIMD partition upon termination of the MIMD program executed by the launcher program. | 11-20-2008 |
20080288747 | Executing Multiple Instructions Multiple Data ('MIMD') Programs on a Single Instruction Multiple Data ('SIMD') Machine - Executing MIMD programs on a SIMD machine, including establishing SIMD partitions on the SIMD machine; booting SIMD partitions in MIMD mode; executing MIMD programs on the compute nodes of a first SIMD partition booted in MIMD mode; re-executing a launcher program by an operating system on a compute node in the first SIMD partition booted in MIMD mode upon termination of the MIMD program executed by the launcher program; determining by a scheduler that the first SIMD partition booted in MIMD mode is required to establish a new SIMD partition large enough to run a SIMD program that is scheduled for execution; moving by the scheduler data processing operations from the first SIMD partition booted in MIMD mode to the second SIMD partition booted in MIMD mode; and establishing by the scheduler the new SIMD partition. | 11-20-2008 |
20090024830 | Executing Multiple Instructions Multiple Data ('MIMD') Programs on a Single Instruction Multiple Data ('SIMD') Machine - Executing Multiple Instructions Multiple Data (‘MIMD’) programs on a Single Instruction Multiple Data (‘SIMD’) machine, the SIMD machine including a plurality of compute nodes, each compute node capable of executing only a single thread of execution, the compute nodes initially configured exclusively for SIMD operations, the SIMD machine further comprising a data communications network, the network comprising synchronous data communications links among the compute nodes, including establishing a SIMD partition comprising a plurality of the compute nodes; booting the SIMD partition in MIMD mode; executing by launcher programs a plurality of MIMD programs on compute nodes in the SIMD partition; and re-executing a launcher program by an operating system on a compute node in the SIMD partition upon termination of the MIMD program executed by the launcher program. | 01-22-2009 |
20090024831 | Executing Multiple Instructions Multiple Data ('MIMD') Programs on a Single Instruction Multiple Data ('SIMD') Machine - Executing MIMD programs on a SIMD machine, including establishing on the SIMD machine a plurality of SIMD partitions; booting a first SIMD partition in MIMD mode; executing, on a compute node of the first SIMD partition booted in MIMD mode, a MIMD accelerator program; executing a SIMD program in a second SIMD partition, one instance of the SIMD program executing on each compute node of the second SIMD partition, each instance of the SIMD program carrying out a portion of the data processing effected by the SIMD program; and accelerating, by an instance of the SIMD program through the MIMD accelerator program, a portion of the data processing of the instance of the SIMD program. | 01-22-2009 |
20090037376 | DATABASE RETRIEVAL WITH A UNIQUE KEY SEARCH ON A PARALLEL COMPUTER SYSTEM - An apparatus and method retrieves a database record from an in-memory database of a parallel computer system using a unique key. The parallel computer system performs a simultaneous search on each node of the computer system using the unique key and then utilizes a global combining network to combine the results from the searches of each node to efficiently and quickly search the entire database. | 02-05-2009 |
20090037377 | DATABASE RETRIEVAL WITH A NON-UNIQUE KEY ON A PARALLEL COMPUTER SYSTEM - An apparatus and method retrieves a database record from an in-memory database of a parallel computer system using a non-unique key. The parallel computer system performs a simultaneous search on each node of the computer system using the non-unique key and then utilizes a global combining network to combine the local results from the searches of each node to efficiently and quickly search the entire database. | 02-05-2009 |
20090037512 | MULTI-NODAL COMPRESSION TECHNIQUES FOR AN IN-MEMORY DATABASE - Embodiments of the invention may be used to distribute a database across a plurality of compute nodes of a parallel computing system. That is, to a method for creating a fully in-memory database on the parallel computing system. Further, data compression techniques may be used increase the performance of the in-memory database by compressing some portions of the database to fit within a single node or a logically or physically related group of nodes. | 02-05-2009 |
20090067334 | MECHANISM FOR PROCESS MIGRATION ON A MASSIVELY PARALLEL COMPUTER - Embodiments off the invention provide a mechanism for process migration on a massively parallel computer system. In particular, embodiments of the invention may be used to update process state data for a migrated compute node, such as MPI (or other communication library) state data, across a full collection of compute nodes present in a given parallel system executing a parallel task. Migrating a process form one compute node to another may be useful to address a variety of sub-optimal operating conditions. For example, one or more processes may be migrated to cure network congestion resulting from a poorly mapped task or when a compute node is predicted to experience a hardware failure. | 03-12-2009 |
20090083276 | INSERTING DATA INTO AN IN-MEMORY DISTRIBUTED NODAL DATABASE - A method and apparatus loads data to an in-memory database across multiple nodes in a parallel computing system. A database loader uses SQL flags, historical information gained from monitoring prior query execution times and patterns, and node and network configuration to determine how to effectively cluster data attributes across multiple nodes. The database loader may also allow a system administrator to force placement of database structures in particular nodes. | 03-26-2009 |
20090083277 | NODAL DATA NORMALIZATION - Embodiments of the invention may be used to normalize data stored in an in-memory database on a parallel computer system. The data normalization may be used to achieve memory savings, thereby reducing the number of compute nodes required to store an in-memory database. Thus, as a result, faster response times may be achieved when querying the data. In one embodiment, normalization may be performed in a manner to avoid datasets that cross physical or logical boundaries of the compute nodes of a parallel system. | 03-26-2009 |
20090089013 | INTERACTIVE TOOL FOR VISUALIZING PERFORMANCE DATA IN REAL-TIME TO ENABLE ADAPTIVE PERFORMANCE OPTIMIZATION AND FEEDBACK - An interactive tool is disclosed for visualizing performance data in real-time to enable adaptive performance optimization for an application running on a massively parallel computer system. The interactive tool may be used to visualize network congestion (and other) performance counters for an application as it runs on the parallel system in real-time. Further, a developer may use the interactive tool to experiment with various tuning options and optimization approaches on-the-fly. | 04-02-2009 |
20090089670 | INTERACTIVE TOOL FOR VISUALIZING PERFORMANCE DATA IN REAL-TIME TO ENABLE ADAPTIVE PERFORMANCE OPTIMIZATION AND FEEDBACK - An interactive tool is disclosed for visualizing performance data in real-time to enable adaptive performance optimization for an application running on a massively parallel computer system. The interactive tool may be used to visualize network congestion (and other) performance counters for an application as it runs on the parallel system in real-time. Further, a developer may use the interactive tool to experiment with various tuning options and optimization approaches on-the-fly. | 04-02-2009 |
20090112953 | ENHANCED GARBAGE COLLECTION IN A MULTI-NODE ENVIRONMENT - Embodiments of the invention enhance a garbage collection process running on a parallel system or distributed computing environment. Using a garbage collector in such an environment allows a more in-depth analysis to be performed than is possible on other systems. This is because the number of compute nodes present in many parallel systems, and the connections between them, allows the overhead of doing advanced analysis to be spread across the nodes and the results of that analysis to be shared among the nodes. | 04-30-2009 |
20090113438 | OPTIMIZATION OF JOB DISTRIBUTION ON A MULTI-NODE COMPUTER SYSTEM - A method and apparatus optimizes job and data distribution on a multi-node computing system. A job scheduler distributes jobs and data to compute nodes according to priority and other resource attributes to ensure the most critical work is done on the nodes that are quickest to access and with less possibility of node communication failure. In a tree network configuration, the job scheduler distributes critical jobs and data to compute nodes that are located closest to the I/O nodes. Other resource attributes include network utilization, constant data state, and class routing. | 04-30-2009 |
20090125611 | SHARING LOADED JAVA CLASSES AMONG A PLURALITY OF NODES - Methods, apparatus, and products are disclosed for sharing loaded Java classes among a plurality of nodes connected together for data communications using a data communication network, the plurality of nodes including an execution node and other nodes, that include: executing, by the execution node, a Java application, including identifying a Java class utilized for the Java application; determining, by the execution node, whether the Java class is already loaded on at least one of the other nodes; retrieving, by the execution node, the loaded Java class from the other nodes if the Java class is already loaded on at least one of the other nodes; and executing, by the execution node, the Java application using the loaded Java class retrieved from the other nodes. | 05-14-2009 |
20090125616 | OPTIMIZED PEER-TO-PEER FILE TRANSFERS ON A MULTI-NODE COMPUTER SYSTEM - A method and apparatus performs peer-to-peer file transfers on a High Performance Computing (HPC) cluster such as a Beowulf cluster. A peer-to-peer file tracker (PPFT) allows operating system, application and data files to be moved from a pre-loaded node to another node of the HPC cluster. A peer-to-peer (PTP) client is loaded into the nodes to facilitate PTP file transfers to reduce loading on networks, network switches and file servers to reduce the time needed to load the nodes with these files to increase overall efficiency of the multi-node computing system. The selection of the nodes participating in file transfers can be based on network topology, network utilization, job status and predicted network/computer utilization. This selection can be dynamic, changing during the file transfers as resource conditions change. The policies used to choose resources can be configured by an administrator. | 05-14-2009 |
20090125883 | NODE SELECTION FOR EXECUTING A JAVA APPLICATION AMONG A PLURALITY OF NODES - Methods, apparatus, and products are disclosed for node selection for executing a Java application among a plurality of nodes connected together for data communications using a data communication network, the plurality of nodes also connected to a service node, that include: tracking, by the service node, loaded Java classes currently loaded on each of the plurality of nodes; receiving, in the service node, an instruction to execute a Java application using the plurality of nodes; identifying, by the service node, Java classes utilized in executing the Java application; selecting, by the service node, one of the plurality of nodes for executing the Java application in dependence upon the loaded Java classes and the Java classes utilized in executing the Java application; and configuring, by the service node, the Java application for execution on the selected node. | 05-14-2009 |
20090132541 | MANAGING DATABASE RESOURCES USED FOR OPTIMIZING QUERY EXECUTION ON A PARALLEL COMPUTER SYSTEM - Embodiments of the invention may be used to increase query processing parallelism of an in-memory database stored on a parallel computing system. A group of compute nodes each store a portion of data as part of the in-memory database. Further, a pool of compute nodes may be reserved to create copies of data from the compute nodes of the in-memory database as part of query processing. When a query is received for execution, the query may be evaluated to determine whether portions of in-memory should be duplicated to allow multiple elements of the query (e.g., multiple query predicates) to be evaluated in parallel. | 05-21-2009 |
20090132609 | REAL TIME DATA REPLICATION FOR QUERY EXECUTION IN A MASSIVELY PARALLEL COMPUTER - Embodiments of the invention may be used to increase query processing parallelism of an in-memory database stored on a parallel computing system. A group of compute nodes each store a portion of data as part of the in-memory database. Further, a pool of compute nodes may be reserved to create copies of data from the compute nodes of the in-memory database as part of query processing. When a query is received for execution, the query may be evaluated to determine whether portions of in-memory should be duplicated to allow multiple elements of the query (e.g., multiple query predicates) to be evaluated in parallel. | 05-21-2009 |
20090144337 | COMMITMENT CONTROL FOR LESS THAN AN ENTIRE RECORD IN AN IN-MEMORY DATABASE IN A PARALLEL COMPUTER SYSTEM - In a networked computer system that includes multiple interconnected nodes, a commitment control mechanism allows designating certain portions of a record in an in-memory database as mandatory and other portions of the record as secondary, and performs mandatory commitment control once all the mandatory portions are available even if one or more secondary portions are not yet available. The secondary portions may be under separate commitment control that is asynchronous to the commitment control for the mandatory portions, or may be under no commitment control at all. The result is a commitment control mechanism that performs commitment control for portions of a record that are marked mandatory even when one or more of the portions marked secondary are not available. | 06-04-2009 |
20090154486 | Tracking Network Contention - Methods, apparatus, and product for tracking network contention on links among compute nodes of an operational group in a point-to-point data communications network of a parallel computer are disclosed. In embodiments of the present invention, each compute node is connected to an adjacent compute node in the point-to-point data communications network through a link. Tracking network contention according to embodiments of the present invention includes maintaining, by a network contention module on each compute node in the operational group, a local contention counter for each compute node, each local contention counter representing network contention on links among the compute nodes originating from the compute node; and maintaining a global contention counter, the global contention counter representing network contention currently on all links among the compute nodes in the operational group. | 06-18-2009 |
20090158276 | DYNAMIC DISTRIBUTION OF NODES ON A MULTI-NODE COMPUTER SYSTEM - A method and apparatus dynamically distribute I/O nodes on a multi-node computing system. An I/O configuration mechanism located in the service node of a multi-node computer system controls the distribution of the I/O nodes. The I/O configuration mechanism uses job information located in a job record to initially configure the I/O node distribution. The I/O configuration mechanism further monitors the I/O performance of the executing job to then dynamically adjusts the I/O node distribution based on the I/O performance of the executing job. | 06-18-2009 |
20090187984 | DATASPACE PROTECTION UTILIZING VIRTUAL PRIVATE NETWORKS ON A MULTI-NODE COMPUTER SYSTEM - A method and apparatus provide data security on a parallel computer system using virtual private networks. An access setup mechanism sets up access control data in the nodes that describes which virtual networks are protected and what applications have access to the protected private networks. When an application accesses data on a protected virtual network, a network access mechanism determines the data is protected and intercepts the data access. The network access mechanism in the kernel may also execute a rule depending on the kind of access that was attempted to the virtual network. Authorized access to the private networks can be made via a system call to the access control mechanism in the kernel. The access control mechanism enforces policy decisions on which data can be distributed through the system via an access control list or other security policies. | 07-23-2009 |
20090240930 | Executing An Application On A Parallel Computer - Methods, systems, and products are disclosed for executing an application on a parallel computer including a plurality of nodes connected together through a data communications network. Each node has a plurality of processors capable of operating independently for serial processing and capable of operating symmetrically for parallel processing. The application has parallel segments for parallel processing and serial segments for serial processing. Embodiments of the invention include: booting up a first subset of the plurality of nodes in a serial processing mode; booting up a second subset of the plurality of nodes in a parallel processing mode; and executing the application on the plurality of nodes, including: migrating the application to the nodes booted up in the parallel processing mode upon encountering the parallel segments during execution, and migrating the application to the nodes booted up in the serial processing mode upon encountering the serial segments during execution. | 09-24-2009 |
20090248894 | Determining A Path For Network Traffic Between Nodes In A Parallel Computer - Determining a path for network traffic between a source compute node and a destination compute node in a parallel computer including identifying a group of compute nodes, the group of compute nodes having topological network locations included in a predefined topological shape; selecting, from the predefined topological shape, in dependence upon a global contention counter stored on the source compute node, a path on which to send a data communications message from the source compute node to the destination compute node; and sending, by the messaging module of the source compute node, the data communications message along the selected path for network traffic between the source and destination compute nodes. | 10-01-2009 |
20090248895 | Determining A Path For Network Traffic Between Nodes In A Parallel Computer - Determining a path for network traffic between a source compute node and a destination compute node in a parallel computer including: beginning with an identified group of compute nodes that includes the source compute node and iteratively until an identified group of compute nodes includes the destination compute node: identifying a group of compute nodes, the group of compute nodes having topological network locations included in a predefined topological shape; selecting a path for network traffic between compute nodes having topological network locations included in the predefined topological shape, and when an identified group of compute nodes includes the destination compute node: selecting a final path for network traffic; and sending a data communications message along the path for network traffic between the source compute node and the destination compute node, the path including, in order of selection, the selected paths and the selected final path. | 10-01-2009 |
20090265305 | Partial Indexes for Multi-Node Database - Embodiments of the invention enable a database spread over multiple nodes to allow each node to have different indexes over the data in tables, depending on how each node would benefit (or not benefit) from having the index(es). When a database table is spread across the nodes of a multi-node or distributed system, each node may maintain only the portion of the index relevant to that node, if doing so would improve the performance of query processing operations on that node. Further, the database may periodically redistributed across the compute nodes based on index performance. Doing so allows the database system to intelligently trade off between consuming space for the index on a node and the usefulness of having an index on that node. | 10-22-2009 |
20090265306 | Index Maintenance in a Multi-Node Database - Embodiments of the invention enable a database spread over multiple nodes to allow each node to have different indexes over the data in tables, depending on how each node would benefit (or not benefit) from having the index(es). When a database table is spread across the nodes of a multi-node or distributed system, each node may maintain only the portion of the index relevant to that node, if doing so would improve the performance of query processing operations on that node. Further, the database may periodically redistributed across the compute nodes based on index performance. Doing so allows the database system to intelligently trade off between consuming space for the index on a node and the usefulness of having an index on that node. | 10-22-2009 |
20090271588 | Migrating Program Objects in a Multi-Node Computer System - Embodiments of the invention enable application programs running across multiple compute nodes of a highly-parallel system to selectively migrate objects from one node to another. For example, when an object becomes too large, a node containing the object may migrate the object to another node, thereby freeing memory space. Whether a large object is migrated can be dependent on how frequently the object is used by the application. Because the memory used by such an object is freed for other uses by the application, overall application performance may be improved. On large parallel systems with thousands of compute nodes, even relatively small improvements in application performance an individual compute node may be magnified many times, resulting in dramatic improvements in overall application performance. | 10-29-2009 |
20090271595 | Configuring An Application For Execution On A Parallel Computer - Methods, systems, and products are disclosed for configuring an application for execution on a parallel computer that include: booting up a first subset of a plurality of nodes in a serial processing mode; booting up a second subset of the plurality of nodes in a parallel processing mode; profiling, prior to application deployment on the parallel computer, the application to identify the serial segments and the parallel segments of the application; and deploying the application for execution on the parallel computer in dependence upon the profile of the application and proximity within the data communications network of the nodes in the first subset relative to the nodes in the second subset. | 10-29-2009 |
20090271596 | Executing An Application On A Parallel Computer - Methods, systems, and products are disclosed for executing an application on a parallel computer having a plurality of nodes. Executing an application on a parallel computer includes: booting up a first subset of a plurality of nodes in a serial processing mode; booting up a second subset of the plurality of nodes in a parallel processing mode; profiling, prior to application execution, an application to identify serial segments of the application, parallel segments of the application, and application data utilized by each of the serial segments and the parallel segments; and executing the application on the plurality of nodes, including migrating, in dependence upon the profile for the application upon encountering the parallel segments during execution, only specific portions of the application and the application data from the nodes booted up in the serial processing mode to the nodes booted up in the parallel processing mode. | 10-29-2009 |
20090271775 | Optimizing Just-In-Time Compiling For A Java Application Executing On A Compute Node - Methods, systems, and products are disclosed for optimizing just-in-time (‘JIT’) compiling for a Java application executing on a compute node, the compute node having installed upon it a Java Virtual Machine (‘JVM’) capable of supporting the Java application, that include: identifying, by an application manager, a particular portion of the Java application; assigning, by the application manager, a JIT level to the particular portion of the Java application; and jitting, by the JVM installed on the compute node, the particular portion of the Java application in dependence upon the JIT level assigned to that particular portion of the Java application. | 10-29-2009 |
20090271807 | Selectively Generating Program Objects on Remote Node of a Multi-Node Computer System - Embodiments of the invention enable application programs running across multiple compute nodes of a highly-parallel system to selectively migrate objects from one node to another. For example, when an object becomes too large, a node containing the object may migrate the object to another node, thereby freeing memory space. Whether a large object is migrated can be dependent on how frequently the object is used by the application. Because the memory used by such an object is freed for other uses by the application, overall application performance may be improved. On large parallel systems with thousands of compute nodes, even relatively small improvements in application performance an individual compute node may be magnified many times, resulting in dramatic improvements in overall application performance. | 10-29-2009 |
20090300752 | UTILIZING VIRTUAL PRIVATE NETWORKS TO PROVIDE OBJECT LEVEL SECURITY ON A MULTI-NODE COMPUTER SYSTEM - The disclosure herein provides data security on a parallel computer system using virtual private networks connecting the nodes of the system. A mechanism sets up access control data in the nodes that describes a number of security classes. Each security class is associated with a virtual network. Each user on the system is associated with one of the security classes. Each database object to be protected is given an attribute of a security class. Database objects are loaded into the system nodes that match the security class of the database object. When a query executes on the system, the query is sent to a particular class or set of classes such that the query is only seen by those nodes that are authorized by the equivalent security class. In this way, the network is used to isolate data from users that do not have proper authorization to access the data. | 12-03-2009 |
20090307287 | Database Journaling in a Multi-Node Environment - A database spread over multiple nodes allows each node to store a journal recording changes made to the database and also allows a journaling component to manage the memory space available for journaling. Two threshold size values may be specified for the journal. The first threshold value specifies a journal size at which to being pruning the journal on a given node. A journal pruning algorithm may be used to identify journal entries that may be removed. For example, once a given transaction completes (i.e., commits) the journal entries related to that transaction may be pruned from the journal. The second threshold value specifies the maximum size of the journal. After reaching this size, journal entries may be written to disk instead of the in-memory journal. | 12-10-2009 |
20090307290 | Database Journaling in a Multi-Node Environment - A database spread over multiple nodes allows each node to store a journal recording changes made to the database and also allows a journaling component to manage the memory space available for journaling. Two threshold size values may be specified for the journal. The first threshold value specifies a journal size at which to being pruning the journal on a given node. A journal pruning algorithm may be used to identify journal entries that may be removed. For example, once a given transaction completes (i.e., commits) the journal entries related to that transaction may be pruned from the journal. The second threshold value specifies the maximum size of the journal. After reaching this size, journal entries may be written to disk instead of the in-memory journal. | 12-10-2009 |
20090307466 | Resource Sharing Techniques in a Parallel Processing Computing System - A method, apparatus, and program product share a resource in a computing system that includes a plurality of computing cores. A request from a second execution context (“EC”) to lock the resource currently locked by a first EC on a first core causes replication of the second EC as a third EC on a third core. The first and third ECs are executed substantially concurrently. When the first EC modifies the resource, the third EC is restarted after the resource has been modified. Alternately, a first EC is configured in a first core and shadowed as a second EC in a second core. In response to a blocked lock request, the first EC is halted and the second EC continues. After granting a lock, it is determined whether a conflict has occurred and the first and second EC are particularly synchronized to each other in response to that determination. | 12-10-2009 |
20090313452 | MANAGEMENT OF PERSISTENT MEMORY IN A MULTI-NODE COMPUTER SYSTEM - A method and apparatus creates and manages persistent memory (PM) in a multi-node computing system. A PM Manager in the service node creates and manages pools of nodes with various sizes of PM. A node manager uses the pools of nodes to load applications to the nodes according to the size of the available PM. The PM Manager can dynamically adjust the size of the PM according to the needs of the applications based on historical use or as determined by a system administrator. The PM Manager works with an operating system kernel on the nodes to provide persistent memory for application data and system metadata. The PM Manager uses the persistent memory to load applications to preserve data from one application to the next. Also, the data preserved in persistent memory may be system metadata such as file system data that will be available to subsequent applications. | 12-17-2009 |
20090319621 | Message Flow Control in a Multi-Node Computer System - Embodiments of the invention provide for controlling message flow across a parallel computer system having multiple compute nodes by selectively grouping compute nodes of such a system into node pools and assigning message flow control policies to nodes in the node pools. The message flow control policies specify logging and/or tracing activities to be performed by instances of applications running on nodes assigned to the node pools. As the application is executed, logging and/or tracing messages are generated on the compute nodes according to message flow control policies assigned to the nodes. Optionally, the message flow is analyzed, the message flow control policies are adjusted, and duplicate messages are eliminated. | 12-24-2009 |
20090320003 | Sharing Compiler Optimizations in a Multi-Node System - Embodiments of the invention enable application programs running across multiple compute nodes of a highly-parallel system to compile source code into native instructions, and subsequently share the optimizations used to compile the source code with other nodes. For example, determining what optimizations to use may consume significant processing power and memory on a node. In cases where multiple nodes exhibit similar characteristics, it is possible that these nodes may use the same set of optimizations when compiling similar pieces of code. Therefore, when one node compiles source code into native instructions, it may share the optimizations used with other similar nodes, thereby removing the burden for the other nodes to figure out which optimizations to use. Thus, while one node may suffer a performance hit for determining the necessary optimizations, other nodes may be saved from this burden by simply using the optimizations provided to them. | 12-24-2009 |
20090320008 | Sharing Compiler Optimizations in a Multi-Node System - Embodiments of the invention enable application programs running across multiple compute nodes of a highly-parallel system to compile source code into native instructions, and subsequently share the optimizations used to compile the source code with other nodes. For example, determining what optimizations to use may consume significant processing power and memory on a node. In cases where multiple nodes exhibit similar characteristics, it is possible that these nodes may use the same set of optimizations when compiling similar pieces of code. Therefore, when one node compiles source code into native instructions, it may share the optimizations used with other similar nodes, thereby removing the burden for the other nodes to figure out which optimizations to use. Thus, while one node may suffer a performance hit for determining the necessary optimizations, other nodes may be saved from this burden by simply using the optimizations provided to them. | 12-24-2009 |
20100185718 | PERFORMING PROCESS MIGRATION WITH ALLREDUCE OPERATIONS - Compute nodes perform allreduce operations that swap processes at nodes. A first allreduce operation generates a first result and uses a first process from a first compute node, a second process from a second compute node, and zeros from other compute nodes. The first compute node replaces the first process with the first result. A second allreduce operation generates a second result and uses the first result from the first compute node, the second process from the second compute node, and zeros from others. The second compute node replaces the second process with the second result, which is the first process. A third allreduce operation generates a third result and uses the first result from first compute node, the second result from the second compute node, and zeros from others. The first compute node replaces the first result with the third result, which is the second process. | 07-22-2010 |
20120144132 | MANAGEMENT OF PERSISTENT MEMORY IN A MULTI-NODE COMPUTER SYSTEM - A method and apparatus creates and manages persistent memory (PM) in a multi-node computing system. A PM Manager in the service node creates and manages pools of nodes with various sizes of PM. A node manager uses the pools of nodes to load applications to the nodes according to the size of the available PM. The PM Manager can dynamically adjust the size of the PM according to the needs of the applications based on historical use or as determined by a system administrator. The PM Manager works with an operating system kernel on the nodes to provide persistent memory for application data and system metadata. The PM Manager uses the persistent memory to load applications to preserve data from one application to the next. Also, the data preserved in persistent memory may be system metadata such as file system data that will be available to subsequent applications. | 06-07-2012 |
20120151573 | UTILIZING VIRTUAL PRIVATE NETWORKS TO PROVIDE OBJECT LEVEL SECURITY ON A MULTI-NODE COMPUTER SYSTEM - The disclosure herein provides data security on a parallel computer system using virtual private networks connecting the nodes of the system. A mechanism sets up access control data in the nodes that describes a number of security classes. Each security class is associated with a virtual network. Each user on the system is associated with one of the security classes. Each database object to be protected is given an attribute of a security class. Database objects are loaded into the system nodes that match the security class of the database object. When a query executes on the system, the query is sent to a particular class or set of classes such that the query is only seen by those nodes that are authorized by the equivalent security class. In this way, the network is used to isolate data from users that do not have proper authorization to access the data. | 06-14-2012 |
20120203763 | DATABASE QUERY OPTIMIZER THAT TAKES NETWORK CHOICE INTO CONSIDERATION - In a networked computer system that includes multiple nodes and multiple networks interconnecting the nodes, a database query optimizer takes network choice into consideration when optimizing a query. The query optimizer logs historical information for queries that includes network information. A network monitor determines network characteristics of the multiple networks. The query optimizer then selects a network for a query based on the network characteristics, and optionally based on the historical information as well. The result is a query optimizer that takes network choice into consideration, thereby providing a query that is optimized according to the characteristics of the networks. | 08-09-2012 |
20130159287 | DATABASE QUERY OPTIMIZER THAT TAKES NETWORK CHOICE INTO CONSIDERATION - In a networked computer system that includes multiple nodes and multiple networks interconnecting the nodes, a database query optimizer takes network choice into consideration when optimizing a query. The query optimizer logs historical information for queries that includes network information. A network monitor determines network characteristics of the multiple networks. The query optimizer then selects a network for a query based on the network characteristics, and optionally based on the historical information as well. The result is a query optimizer that takes network choice into consideration, thereby providing a query that is optimized according to the characteristics of the networks. | 06-20-2013 |
20130191504 | OPTIMIZED PEER-TO-PEER FILE TRANSFERS ON A MULTI-NODE COMPUTER SYSTEM - A method and apparatus performs peer-to-peer file transfers on a High Performance Computing (HPC) cluster such as a Beowulf cluster. A peer-to-peer file tracker (PPFT) allows operating system, application and data files to be moved from a pre-loaded node to another node of the HPC cluster. A peer-to-peer (PTP) client is loaded into the nodes to facilitate PTP file transfers to reduce loading on networks, network switches and file servers to reduce the time needed to load the nodes with these files to increase overall efficiency of the multi-node computing system. The selection of the nodes participating in file transfers can be based on network topology, network utilization, job status and predicted network/computer utilization. This selection can be dynamic, changing during the file transfers as resource conditions change. The policies used to choose resources can be configured by an administrator. | 07-25-2013 |
20130198343 | OPTIMIZED PEER-TO-PEER FILE TRANSFERS ON A MULTI-NODE COMPUTER SYSTEM - A method and apparatus performs peer-to-peer file transfers on a High Performance Computing (HPC) cluster such as a Beowulf cluster. A peer-to-peer file tracker (PPFT) allows operating system, application and data files to be moved from a pre-loaded node to another node of the HPC cluster. A peer-to-peer (PTP) client is loaded into the nodes to facilitate PTP file transfers to reduce loading on networks, network switches and file servers to reduce the time needed to load the nodes with these files to increase overall efficiency of the multi-node computing system. The selection of the nodes participating in file transfers can be based on network topology, network utilization, job status and predicted network/computer utilization. This selection can be dynamic, changing during the file transfers as resource conditions change. The policies used to choose resources can be configured by an administrator. | 08-01-2013 |