International Business Machines Corporation

Globalfoundries Inc.

Vijayshankar Raman

Ronald Jason Barber

Richard Sefton Sidle

Guy Maring Lohman

Gopi K Attaluri

Ippokratis Pandis

Lin Qiao

Adam J Storm

Liping Zhang

Sam Sampson Lightstone

Rene Mueller

Chandrasekaran Mohan

Pinar Tozun

Naresh K Chainani

Vincent Kulandaisamy

Mir Hamid Pirahesh

Bishwaranjan Bhattacharjee

Mohammad Sadoghi Hamedani

Ronen Grosman

David C Sharpe

Garret Frederick Swart

Eugene Jon Shekita

Xun Xue

Volker Gerhard Markl

Yinan Li

Min-Soo Kim

Knut Stolze

Inderpal Singh Narang

Christian Marcelo Garcia-Arellano

Oliver Draese

Yuanyuan Tian

Frederick Ralph Reiss

Peter Haas

Peter Bendel

Gregory R Stager

Wayne Jonleon Young

Stefan Arndt

Viktor Leis

F Ryan Johnson

Matthew Albert Huras

Ian R Finlay

Bruce Gilbert Lindsay

David Everett Simmen

Jens Leenstra

Wei Han

Peter Martin Schwarz

Robert S Germain

Kiran Chinta

Holger Kache

Ivan Popivanov

Ashraf Aboulnaga

Jae Gil Lee

Stephan Eberhard Ewen

Kavitha Ranganathan

Vincent Kulandai Samy

Eduard Diner

Marki Volker

Harry M Yudenfriend

David Darden Chambliss

Namik Hrle

Felix O Beier

You-Chin Gene Fuh

James Alan Ruddy

Fen-Ling Lin

Oliver Koeth

Prashant Pandey

Guillermo Alvarez

Sandeep Tata

Robert Krauthgamer

Chen Luo

Tianchao Li

Matthew J Spilchen

Francisco J Bermudez

Atri Rudra

Ying-Lin Chen

Kouhorng Allen Yang

Pablo Diaz-Gutierrez

Carlos J Caballero Grolimund

Aranyak Mehta

Ioana M Delaney

A method includes storing an anchor row vector identification for an anchor row to a local memory. It is determined whether the anchor row vector identification is visible based on isolation requirements. The anchor row vector identification is accessed upon a determination that the anchor row vector identification is visible, and the row vector identification is re-read from the local memory. It is determined whether the anchor row vector identification has not changed since a start of the accessing. Upon a determination that the anchor row vector identification has not changed, read anchor row fields are returned. A first check history is performed on an anchor row history tuple sequence number (TSN) for the anchor row.

In-place updates with concurrent reads in a decomposed state

One embodiment provides for a method to track replication state and providing quorum visible retrievals. The method includes assigning, by a writer node performing write operations, a sequence number. The sequence number is replicated with a write operation to other replica nodes in an asynchronous fashion thru a first network channel. On a second network channel, a state of a local system is communicated by sending a latest sequence number available that has been hardened to the other replica nodes. A query is obtained by a reader node, and a sequence number limit is selected. Results up to the sequence number limit are returned.

Tracking data availability using heartbeats

Embodiments relate to aggregation operations. More specifically, embodiments relate to enforcing an aggregation constraint across a distributed network of compute nodes responding asynchronously. Each compute node individually processes the local partial aggregate value in parallel and asynchronously communicates with a conflict resolution node (CRN).

Scalable enforcement of aggregation constraints within transactions

One embodiment provides for a method including performing, by a processing thread, a process that analyzes transactional operations by maintaining the transactional operations in transaction local side logs, and waiting until a successful transaction commit to append the transaction local side logs to a log stream. The processing thread processes the transactional operations on a key used to determine whether existing data is found for the key. The transactional operations are sped up through parallelism based on partitioning tables across nodes handling the transactional operations. A first process is performed by a first processor that processes updates for values of a key based on updating a first start time table index using unique keys and a start time field of a row for a first appearance of each unique key from the transactional operations.

Resolving versions in an append-only large-scale data store in distributed data management systems

A first table has more than one key column, and a hash array is built for the table, which includes computing hash indices for respective rows of the first table based on more than one key column. Building the hash array also includes writing in the hash array for each row at a respective location in the hash array, wherein each location is indicated by each row's respective hash index computed from the more than one key column. The hash array is probed responsive to the second table. In the building of the hash array for the first table, the writing in the hash array for each row at a respective location in the hash array writes only one of the key columns as a single key field in the hash array for each row.

Technology to reduce cost of concatenation for hash array

Embodiments of the invention relate to processing data records, and for a multi-phase partitioned data reduction. The first phase relates to processing data records and partitioning the records into a first partition of records having a common characteristic and a second partition of records that are not members of the first partition. The data records in each partition are subject to intra-partition data reduction responsive to a resource constraint. The data records in each partition are also subject to an inter-partition data reduction, also referred to as an aggregation to reduce a footprint for storing the records. Partitions and/or individual records are logically aggregated and a data reduction operation for the logical aggregation of records takes place in response to available resources.

Accurate partition sizing for memory efficient reduction operations

One embodiment provides for monotonic transactions in a multi-master database with loosely coupled nodes including performing, by a processor, a read transaction protocol including: performing, by the processor, a read transaction protocol including issuing read transactions at any one of the nodes; attempting to return rows that are known to be replicated by at least a read quorum of the nodes; and for timing-out of the attempting to return rows known to be replicated by at least the read quorum of the nodes based on the special-node being non-responsive: selecting a new special-node from the nodes; and repeating attempting to return rows that are known to be replicated by at least the read quorum of the nodes.

Monotonic transactions in a multi-master database with loosely coupled nodes

According to one or more embodiments of the present invention, a computer-implemented method includes creating a hybrid index for a multi-zone data storage system in which a first zone has data that is organized for transaction processing, and a second zone that has data organized for analytic processing. Each index entry in the hybrid index includes an index key including an equality column, a begin-timestamp column, and a tuple-sequence-number column. The method further includes building an index run to be included in the hybrid index, the index run includes multiple index entries. An index run is created by selecting columns by scanning data blocks from the first zone, computing tuple-sequence-number values for the index entries to be included in the hybrid index, and sorting the index entries according to a predetermined order using the columns to be included in the index run.

Indexing for evolving large-scale datasets in multi-master hybrid transactional and analytical processing systems

A computer-implemented method for processing a transaction may include, upon receiving a transaction change request, operating a coordinator to generate a unique transaction identifier for the transaction request, send a data change request along with the unique transaction identifier to sources, and store the unique transaction identifier along with a respective first change count in a transaction status table. The method may also include, upon receiving a query, operating a given source to determine from the transaction status table, a second change count of sources associated with the unique transaction identifier, determine whether the second change count matches the first change count for each source, and when the first and second change counts match in each of the other sources, read data associated with the query from the given source and each source.

Count and transaction identifier based transaction processing

One embodiment provides for a method for grooming transactional operations in a multi-master distributed data management system including maintaining transactional operations in at least one transaction-local side-log. A table log scan is performed for a table including recording identification for each record in a log. Tuple sequence numbers are generated for the records and the records are written on groomed transaction blocks to the table.

Transactional operations in multi-master distributed data management systems

A method includes performing, by a data structure processor, concurrent read and write operations into a hierarchical data structure that includes a mutable tier including extendible hashing, a hash table, and an immutable tier including a concise hash table (CHT) bitmap. Writers acquire latches on the hierarchical data structure elements that the latches modify. The hierarchical data structure elements are directly accessed by readers without acquiring latches. A modify operation is executed by a writer for one or more levels of the hierarchical data structure. When removed portions of the hierarchical data structure are no longer referenced, tracking is performed by use of a combination of a global state value and a copied local state value.

Concurrent reads and inserts into a data structure without latching or waiting by readers

A method creating a compact hash table includes a thread executing using a processor for building a linear probe hash table based on allocating space for an array of keys. Payloads are inserted into the linear probe hash table. Keys from the array of keys are inserted into the linear probe hash table with linear probing. A linear probing bitmap is generated based on occupied locations in the linear probe hash table. A compact hash table is created based on the linear probe hash table and the linear probing bitmap.

Parallel build of non-partitioned join hash tables and non-enforced N:1 join hash tables

One embodiment provides for a method including processing transactional operations on a key used to determine whether existing data is found for that key. A first time index is updated using unique keys and a start time field of a first appearance of each key from the transactional operations. A deferred update of prior versions of the key is performed for non-recent data upon determining that recent data in the transactional operations is found for the key.

One embodiment provides a method for efficient representation, access and modification of variable length data objects including converting, by a processor, all object updates into appends. The processor divides each object update into bounded size tokens. The tokens are stored normalized as an anchor and a linked list of fragments, the anchor having a pointer to a linked list of fragments. A fragment comprises a contiguous list of tokens that correspond to the appends.

Efficient representation, access and modification of variable length objects

A method includes performing a lookup using a key into a root node of a multi-tier data structure, to find a partition for performing an insert. A lookup for the key is performed on a first level index that is part of a linked data structure. A payload or reference is added to the linked data structure based on data structure criterion, otherwise the key and the payload are added to the linked data structure if the key is not found. A new first level index is created and added to the linked data structure upon the linked data structure remaining unchanged. The key and the payload or reference are added to the new index. Based on merge criterion, a new second level index is created and a portion of content from selected first level and second level indexes are merged for combining into the new second level index.

Fast multi-tier indexing supporting dynamic update

Growing community of inventors

Sunnyvale, CA, United States of America

Vijayshankar Raman