US 12,450,290 B2
Estimating graph size and memory consumption of distributed graph for efficient resource management
Jonas Schweizer, Schlieren (CH); Arnaud Delamare, Zurich (CH); Jinsu Lee, San Mateo, CA (US); Sungpack Hong, Palo Alto, CA (US); Hassan Chafi, Palo Alto, CA (US); and Vasileios Trigonakis, Zurich (CH)
Assigned to Oracle International Corporation, RedwoodShores, CA (US)
Filed by Oracle International Corporation, Redwood Shores, CA (US)
Filed on Oct. 26, 2023, as Appl. No. 18/384,248.
Prior Publication US 2025/0139163 A1, May 1, 2025
Int. Cl. G06F 16/901 (2019.01); G06F 11/32 (2006.01)
CPC G06F 16/9024 (2019.01) [G06F 11/328 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
for a graph to be loaded into memory of one or more machines of a distributed graph processing system, generating a memory consumption estimate by:
sampling graph data of the graph from a source;
estimating graph statistics for the graph based on the sampled graph data;
predicting, based on the graph statistics, an estimated final graph size indicating an amount of memory used to keep a plurality of graph structures in memory,
wherein:
the graph comprises a plurality of entities and one or more edges between entities,
the plurality of graph structures represent the graph, and
the plurality of graph structures comprise at least one vertex data structure representing the plurality of entities of the graph and at least one edge data structure representing the one or more edges between entities; and
predicting, based on the graph statistics, an estimated peak memory usage indicating an upper bound of memory usage across the one or more machines during loading of the graph,
wherein the memory consumption estimate comprises the estimated final graph size and the estimated peak memory usage;
wherein the method is performed by one or more computing devices.