International Business Machines Corporation

Hebert Walter Pereyra

Xun Xue

David C Kalmuk

Jon A Lind

Jack Hon Wai Ng

Calisto Paul Zuzarte

Jeffrey James Goss

Kaarel Truuvert

Roger C Raphael

Paul Miller Bird

Yuk Kuen Chan

Michael Joshua Snowbell

Tim J Vincent

Peter K L Shum

Josep L Lariba-Pey

Catharina K Wong

Josep Aguilar-Saborit

Victor Muntes-Mulero

Josep Aguilar Saborit

Shantanu Sadanand Mundkur

Yeung-Kuen See

Josep L Larriba-Pey

Jesse Lee

Maxim Neaga

Yaxian Wang

Victor Muntés-Mulero

A method, apparatus, system, and computer program code for policy-based enforcement in a data virtualization system is provided. Responsive to receiving a query, a computer identifies a virtual object among a set of connected objects that is represented by a set of data assets and their hierarchical relationships. The virtual object corresponds to a subset of the data assets. The computer identifies a subset of objects according to a cumulative transitive closure of the virtual object over the set of connected objects. The computer identifies a set of policies for the subset of objects. For each object in the subset of objects, the computer determines an intermediate decision according to set of policies, whereby a set of intermediate decisions is formed. The computer deterministically reconciles the set of intermediate decisions to generate a resolved decision. The computer provides access to the queried virtual objects based on the resolved decision.

Deterministic enforcement in data virtualization systems

A data processing system, and an article of manufacturing, join rows associated with a source table column with rows associated with a target table column. A source node and a target node contain the source and target tables, respectively. A reduced representation of selected rows associated with the source table column is generated, as is a representation of the target table column. A filtering module filters the generated reduced representation of selected rows associated with the source table column through the generated representation of the target table column, the filtered generated reduced representation of selected rows identifying source table rows that do not have to be joined with the target table. The rows associated with the source table column minus the filtered generated reduced representation of selected rows are joined to the rows associated with the target table column.

System and program for joining source table rows with target table rows

A database management system in which a plurality of nodes form a database instance, each node including a communication manager for dynamically configuring inter-nodal communication resources. The communication manager receives communication resource allocation requests from clients or a self-tuning algorithm. A resource self-tuning mechanism allocates or de-allocates memory blocks used for communication resource elements dynamically in real time without cycling the instance. Memory blocks are de-allocated asynchronously by placing associated communication resource elements in quarantine until all communication resource elements associated with the memory block are quarantined.

Dynamic configuration and self-tuning of inter-nodal communication resources in a database management system

An operating system directed to using special properties of a common inter-process communications mechanism (IPC), namely UNIX domain socket-pairs or stream-pipes alternatively as a storage medium for file-descriptors of UNIX processes. When a file-descriptor is written into a socket-pair, and closed in the UNIX process, the file remains open, but occupies no space in the process' file-table. The file-descriptor may later be read out of the socket-pair to reestablish it in the file-table, and access it. This property is implemented in an IPC mechanism of UNIX operating system whereby a process such as a dispatcher may manage more connections and processes than its file-table size allow. This provides scalability improvements of the UNIX operating system.

Method and system for optimizing file table usage

A method for dynamically managing memory to support one or more processes executing in a remote direct memory access (RDMA) environment is provided. The method includes inserting a descriptor in a shared descriptor table, the descriptor corresponding to a block of memory allocated to a heap by an operating system. The method further includes, in response to allocating a portion of the block of memory from the heap to a process, determining whether the process has an existing registration with an application program interface for the block of memory. If the process has no existing registration, registering the process the process is registered with the application program interface and a registration corresponding to the block of memory is stored in a private registration table of the process. Additionally, the method includes, in response to the process releasing the allocated portion of the block of memory to the operating system, de-registering with the application program interface and removing the registration from the registration table. When the block of memory is released to the operating system, other registrations corresponding to the block of memory in other private registration tables of other processes remain in the other registration tables.

Dynamic memory management in an RDMA context

Disclosed is a data processing system implemented method, a data processing system, and an article of manufacturing for use with the data processing system. The data processing system implemented method is used for directing the data processing system to join rows associated with a column of a source table with rows associated with a column of a target table. The data processing system includes a source node containing the source table and including a target node containing the target table. The method includes generating a reduced representation of selected rows associated with the column of the source table, and generating a representation of the column of the target table, filtering the generated reduced representation of selected rows associated with the column of the source table through the generated representation of the column of the target table, the filtered generated reduced representation of selected rows identifying source table rows that do not have to be joined with the target table, and joining, to the rows associated with the column of the target table, the rows associated with the column of the source table minus the filtered generated reduced representation of selected rows.

Method, system and program for joining source table rows with target table rows

A communications multiplexor includes dispatcher processes for monitoring client connections. The dispatcher processes detect activity on those connections, and then pass active physical (client) connections to agent processes for servicing. Transfer is done through specific connection queues that are associated with a set of agents. A multi-queuing structure permits pooling of agents on a set of shared resources thereby reducing time required to switch between different client connections. After an agent has serviced a given connection, the agent returns that connection to the agent's dispatcher (there is a static assignment between connections and dispatchers), and then reads the next unit of work from the agent's associated connection queue. This structure may be scalable while allowing optimal performance when passing physical connections between processes.

Communication multiplexor using listener process to detect newly active client connections and passes to dispatcher processes for handling the connections

A gateway provides communication of tightly coupled XA transactions to a server hosting a DBMS and makes efficient use of domain socket pairs to implement demultiplexing in the gateway. The gateway includes a TCP/IP listener process which accepts connection requests from a transaction manager. Logical agents represent the application connection and are managed by a logical agent scheduler and are passed within the gateway using domain socket pairs. Gateway agents are dedicated to identified tightly coupled XA transactions to provide tightly coupled XA transaction support for DBMS systems which do not provide such support directly. The domain socket pairs in the gateway are assigned to logical agent schedulers and are available in a domain socket pool. A wait queue is provided to buffer logical agents when connection to the server is not available or where there are no domain socket pairs available in the domain socket pool. The wait queue is itself implemented as a domain socket pair.

Efficient use of domain socket pairs in communication for tightly coupled transactions

A multi-node relational database management system is provided that can be embodied in software for managing a database having declared temporary tables associated with respective software application programs. The system comprises a plurality of nodes. The nodes include a catalog node for maintaining a list of identifiers for declared temporary tables associated with any application; a coordinator node for an application; at least one subordinate node for the application, the subordinate node having at least one partition of the declared temporary tables of the application; and a declared temporary table manager for causing the coordinator node to maintain for the application a list of declared temporary tables.

Method and apparatus for synchronizing not-logged application temporary tables in a multi-node relational database management system

Growing community of inventors

Toronto, Canada

Hebert Walter Pereyra