Sap Ag

Sap Se

Workday, Inc.

Risk Management Solutions, Inc.

International Business Machines Corporation

Christof Bornhoevd

Horst Werner

Brian S Mo

Wolfgang Lehner

Matthias M Wiemann

Marcus Paradies

Darren Kermit Lee

Michael Rudolf

Artin Avanes

Holger Robert Ziekow

Stephan Haller

Rama Krishna Gurram

Asuman Suenbuel

Pablo E Guerrero

Mir Hamid Pirahesh

Chandrasekaran Mohan

Saileshwar Krishnamurthy

Daniel Buchmann

Berthold Reinwald

Srinivas Vemuri

Thomas Mueller

Ashish Patel

Karl Fuerst

Zoltan Nochta

Neil Thombre

Uwe Kubach

Luciana Moreira Sa De Souza

Thomas Seufert

Patrik Spiess

Francis Wang

Cory Isaacson

Thomas Odenwald

Ngoc Nguyen

Brian Mo

David Glaubman

Ilan Cohen

Mehmet Altinel

William Gardella

Zorica Caklovic

Dewei Sun

Nanhong Ye

Curtis Huang

Aabhas Paliwal

Nabil Adam

Basit Shafiq

Mariano A Cilia

Philipp C K Sachs

Alejandro P Buchmann

Dan Lynn

Sumeet Bajaj

Ian McCormick

Crichton McCutcheon

Ramabrahmam Venkata Velury

Yuh-Ming Shyy

Matthias Wiemann

Yunfeng Fei

Aaron Kaminsky

Elena Vasilyeva

Ragy Eleish

Mark Thompson

Maik Thiele

Techniques for providing horizontally scaled caching of versioned data are provided. In some aspects, the techniques described herein relate to a method including initializing a first version cache (VC) object based on a version of data stored in a data storage device; replicating the first VC to generate a second VC; receiving a write operation at the first VC; generating a delta for the write operation, the delta representing a change in the version of data; writing the delta to a persistent replication log, the persistent replication log storing an ordered set of deltas including the delta; writing data in the write operation to the data storage device; and applying the ordered set of deltas at the second VC to update data stored by the second VC.

Horizontal scaling of version caches in a distributed high-concurrency multi-user environment

A system, method, and device for updating a base version of a hypercube are provided. The method includes (i) obtaining a request for a planning session with respect to a hypercube; and (ii) in response to obtaining the request: (a) determining that the planning session uses a base version of the hypercube, (b) receiving a value input to update a particular location in the hypercube, (c) storing the value for the particular location in the base version of the hypercube, and (d) propagating an update to the particular location of the base version of the hypercube to one or more scenario versions having a dependency on the base version.

What if scenario planning—updating a base plan version

The present application discloses a method, system, and computer system for processing queries to sparsely populated datasets. The method includes (i) receiving a query from a client system, (ii) determining, based at least in part on the query, a plurality of partitions of a hypercube that are expected to comprise information used in generating a response for the query, (iii) communicating a set of requests to a data source storing the hypercube, the set of requests comprising one or more requests corresponding to the plurality of partitions of the hypercube, (iv) obtaining, from the data source, a set of results to the set of requests, the set of results comprising data stored in the plurality of partitions of the hypercube, (v) determining a plurality of virtual objects for the set of results, (vi) combining the plurality of virtual objects into a response to the query, and (vii) providing, to the client system, the response to the query.

System and algorithms for fast and scalable data access to high dimensionality with sparse data

Systems and methods for precomputing data and storing cache objects corresponding to the precomputed data are described. A system creates a new cache object when a user interacts with the system. The system precomputes formulas in the newly created cache object by replacing the formulas with corresponding calculated values. The system precomputes the formulas in the background (i.e., the user is not presented with the precomputed values while the user is manipulating the data). The system may persistently store a precomputed version cache object in a dedicated version cache storage for later use. If updates are performed to the structure and/or values of a version represented in a precomputed version cache object, effected parts of the version cache object are invalidated by replacing calculated values with the underlying formulas.

Systems and methods for accelerating data computation

A system for assigning a workload to compute resources includes an interface and a processor. The interface is configured to receive a workload. The processor is configured to break the workload into a set of subproblems; and for a subproblem of the set of subproblems: determine whether the subproblem benefits from intersheet parallelism; determine whether the subproblem benefits from intrasheet parallelism; determine whether the subproblem benefits from directed acyclic graph (DAG) partitioning; and assign the subproblem, wherein assigning the subproblem utilizes optimization when appropriate based at least in part on benefits from the intersheet parallelism, the intrasheet parallelism, and the DAG partitioning.

Systems and methods for improving computational speed of planning by enabling interactive processing in hypercubes

In an embodiment, a method comprises storing metadata that maps a domain model to data stored in a plurality of data stores, each data store being associated with a particular query language, the domain model describing the data and relationships between the data. The method comprises receiving a request for data stored in a first data store of, the request being in a request query language based on the domain model. The method comprises generating an abstract syntax tree indicating a field selection, an entity path, and a condition based on the request; generating a structure model comprising one or more aggregation levels for one or more entities; and generating annotations comprising query language aliases for portions of the request and correlating the portions of the request with the metadata. The method comprises generating queries in a first query language associated with the first data store based on the annotations; and sending the queries to the first data store.

Metadata-based general request translator for distributed computer systems

In an embodiment, a method comprises generating and storing metadata that maps a domain model to data stored in a data store, wherein the domain model describes the data and relationships between the data; receiving, from a client application, a request for data stored in the data store, wherein the request is in a first query language; translating the request into one or more queries based on the metadata, wherein the one or more queries are in a second query language; sending the one or more queries to the data store; receiving a first result set from the data store, based on the data store executing the one or more queries.

A computing system and platform uses various types of data stores to allow efficient querying of, and accelerated access to, extremely large data sets. One such data store is a dimension store that combines key-value and columnar stores, access to which is provided by several selectable mechanisms chosen based the nature of the data of interest. These include bitmap-based access, us of an optimized columnar data format, and access via namespace identifiers. A compressed, optimized page data format is provided for storing and analyzing large fact-based data. The complex dimension store is used to provide complex relationships and interpretation of the fact-based data, enabling high-performance advanced queries, with bitmap indexes passed between the two stores. Dimension data is stored in an encrypted manner throughout the system, and can be exchanged among parties in a secure manner.

High performance big data computing system and platform

A system, computer-implemented method, and computer-readable storage medium for generating a block-based index, are provided. A block index is generated where the block index comprises a plurality of blocks and a block corresponds to a section of a graph column that stores a value. A block range vector is also generated for the index where the block range vector includes range information for the block that corresponds to the section of the graph and where the block-based index facilitates traversal of the graph column that searches for the value by constraining the traversal to the section of the graph.

Index structure to accelerate graph traversal

Among other disclosed subject matter, a computer program product is tangibly embodied in a computer-readable storage medium and includes instructions that when executed by a processor perform a method for recording information about an item. The method includes assigning a unique identifier to an item in a repository, the item supporting versioning and being subject to a data model having no predefined schema for items. The method includes associating a first attribute with the unique identifier, the first attribute being part of the item and being configured for modification. The method includes associating the first attribute with a version number, wherein the versioning is supported without recording in the repository a time period when the first attribute is valid.

Recording information about an item

A system, computer-implemented method, and a computer-readable storage medium for a data graph traversal are provided. The input parameters for traversing the data graph are received. The data graph having a set of vertices and a set of edges are stored in a column based format in a memory cache of a computer device based on the input parameters is traversed. The traversal generates a set of traversed vertices that are the result of the graph traversal.

Graph traversal operator inside a column store

A system, computer-implemented method, and a computer-readable storage medium for a traversal of a property graph, are provided. The edge table of the property graph is divided into a plurality of fragments. A first fragment is selected for traversal using a set of selected vertices, where the traversal identifies a set of edges. Based on the set of edges, a set of adjacent vertices is determined during the traversal. A set of discovered vertices in the property graph is determined based on the set of selected vertices and the set of adjacent vertices.

Graph traversal operator and extensible framework inside a column store

An execution framework allows developers to write sequential computational logic, constrained for the runtime system to efficiently parallelize execution of custom business logic. The framework can be leveraged to overcome limitations in executing low level procedural code, by empowering the system runtime environment to parallelize this code. Embodiments employ algorithmic skeletons in the realm of optimizing/executing data flow graphs of database management systems. By providing an extensible set of algorithmic skeletons the developer of custom logic can select the skeleton appropriate for new custom logic, and then fill in the corresponding computation logic according to the structural template of the skeleton. The skeleton provides a set of constraints known to the execution environment, that can be leveraged by the optimizer and the execution environment to generate parallel optimized execution plans containing custom logic, without the developer having to explicitly describe parallelization of the logic.

Parallel programming of in memory database utilizing extensible skeletons

A system, computer-implemented method, and a computer-readable medium for determining why a query returns a null set in a database management system. When the database management system receives a query graph, database management system compares the query graph to the data graph. Based on the comparison, database management system identifies a discovered component of the query graph and a missing component of the query graph, wherein the missing query component indicates a reason for the null set.

Processing diff-queries on property graphs

Growing community of inventors

Belmont, CA, United States of America

Christof Bornhoevd