Oracle International Corporation

Jinsu Lee

Hassan Chafi

Sungpack Hong

Thomas Manhardt

Nicholas Roth

Vasileios Trigonakis

Siegfried Depner

Arnaud Delamare

Petr Koupy

Vlad Ioan Haprian

Hugo Kapp

Ioannis Alagiannis

Stefan Kaestle

Jonas Schweizer

Ayoub Berdai

Luigi Fusco

Abderrahmane Melhaoui

The present disclosure relates to various approaches for fast and scalable TOP K SHORTEST and CHEAPEST graph queries supporting horizontal aggregations on the group variables of a path in a distributed graph query engine. A distributed graph query processing engine may execute a graph query in a plurality of computing devices. A plurality of subjobs may be generated based at least in part on the graph query. Execution of an asynchronous pattern matching subjob of the plurality of subjobs may be initiated, and, in response to the asynchronous pattern matching subjob identifying one or more source vertices of a plurality of vertices, the execution of the asynchronous pattern matching subjob may be paused. An output context set comprising the one or more source vertices may be generated. Execution of a synchronous path matching subjob of the plurality of subjobs may be initiated, and a reachability map may be generated based at least in part on one or more matched paths between the one or more source vertices and one or more destination vertices of the plurality of vertices. The execution of the asynchronous pattern matching subjob may be resumed based at least in part on the output context set and the reachability map.

Shortest and cheapest paths in distributed asynchronous graph traversals

An estimator is provided that can be used to get an estimate of final graph size and peak memory usage of the graph during loading, based on sampling of the graph data and using machine learning (ML) techniques. A data sampler samples the data from files or databases and estimates some statistics about the final graph. The sampler also samples some information about property data. Given the sampled statistics gathered and estimated by the data sampler, a graph size estimator estimates how much memory is required by the graph processing engine to load the graph. The final graph size represents how much memory will be used to keep the final graph structures in memory once loading is completed. The peak memory usage represents the memory usage upper bound that is reached by the graph processing engine during loading.

Estimating graph size and memory consumption of distributed graph for efficient resource management

In an embodiment, multiple computers cooperate to retrieve content from tables in a relational database. Each table contains respective rows. Each row contains a vertex of a graph. Many high-degree vertices are identified. Each high-degree vertex is connected to respective edges in the graph. A count of the edges of each high-degree vertex exceeds a degree threshold. A central computer detects that all vertices in a high-degree subset of tables are high-degree vertices. Based on detecting the high-degree subset of tables, multiple vertices of the graph that are not in the high-degree subset of tables are replicated. Within local storage capacity limits of the computers, this degree-based replication may be supplemented with other vertex replication strategies that are schema based, content based, or workload based. This intelligent selective replication maximizes system throughput by minimizing graph data access latency based on data locality.

Access-frequency-based entity replication techniques for distributed property graphs with schema

Techniques are described herein for evaluating graph processing tasks using a multi-stage pipelining communication mechanism. In a multi-node system comprising a plurality of nodes, each node of said plurality of nodes executes a respective communication agent object. The respective communication agent object comprises: a sender lambda function is configured to perform sending operations and generate source messages based on the sender operations. An intermediate lambda function is configured to read source messages marked for a node, perform intermediate operations based on the source messages and generate intermediate messages based on the intermediate operations. A final receiver lambda function configured to: read intermediate messages marked for said each node, perform final operations based on the intermediate messages and generate a final result based on the final operations.

Multi-stage pipelining for distributed graph processing

Techniques are provided for dynamically self-balancing communication and computation. In an embodiment, each partition of application data is stored on a respective computer of a cluster. The application is divided into distributed jobs, each of which corresponds to a partition. Each distributed job is hosted on the computer that hosts the corresponding data partition. Each computer divides its distributed job into computation tasks. Each computer has a pool of threads that execute the computation tasks. During execution, one computer receives a data access request from another computer. The data access request is executed by a thread of the pool. Threads of the pool are bimodal and may be repurposed between communication and computation, depending on workload. Each computer individually detects completion of its computation tasks. Each computer informs a central computer that its distributed job has finished. The central computer detects when all distributed jobs of the application have terminated.

Concurrent distributed graph processing system with self-balance

Techniques are described herein for asynchronous execution of queries on statically replicated graph data. In an embodiment, a graph is partitioned among a plurality of computers executing the graph querying engine. One or more high-degree vertices of the graph are each replicated in each graph partition. The partitions, including the replicated high-degree vertices, are loaded in memory of the plurality of computers. To execute a query, a query plan is generated based on the query. The query plan specifies a plurality of operators and an order for the plurality of operators. The order is such that if an operator requires data generated by another operator, then the other operator is ordered before the operator in the query plan. Replicated copies of a vertex is visited if matches made by subsequent operator(s) are limited by data unique to the replicated vertices.

Fast distributed graph query engine

Techniques herein provide job control and synchronization of distributed graph-processing jobs. In an embodiment, a computer system maintains an input queue of graph processing jobs. In response to de-queuing a graph processing job, a master thread partitions the graph processing job into distributed jobs. Each distributed job has a sequence of processing phases. The master thread sends each distributed job to a distributed processor. Each distributed job executes a first processing phase of its sequence of processing phases. To the master thread, the distributed job announces completion of its first processing phase. The master thread detects that all distributed jobs have announced finishing their first processing phase. The master thread broadcasts a notification to the distributed jobs that indicates that all distributed jobs have finished their first processing phase. Receiving that notification causes the distributed jobs to execute their second processing phase. Queues and barriers provide for faults and cancellation.

Distributed graph processing system featuring interactive remote control mechanism including task cancellation

Techniques minimize communication while loading a graph. In a distributed embodiment, each computer loads some edges of the graph. Each edge connects a source vertex (SV) to a destination vertex. For each SV of the edges, the computer hashes the SV to detect a tracking computer (TrC) that tracks on which computer does the SV reside. Each computer informs the TrC that the SV originates an edge that resides on that computer. For each SV, the TrC detects that the SV originates edges that reside on multiple providing computers (PCs). The TrC selects a target computer (TaC) from the multiple PCs to host the SV. The TrC instructs each PC, excluding the TaC, to transfer the SV and related edges that are connected to the SV to the TaC. A vertex's internal identifier indicates which computer hosts the vertex. The TrC maintains a mapping between external and internal identifiers.

Growing community of inventors

San Mateo, CA, United States of America

Jinsu Lee