Granted Patent B2

US 12,541,492 · App. 18/643,242 · Granted Feb 3, 2026

Inventors: George Kondiles (Chicago, IL); Rhett Colin Starr (Long Grove, IL); Joseph Jablonski (Chicago, IL); S. Christopher Gladwin (Chicago, IL)

Assignee: Ocient Inc.

G06F16/221G06F16/2365G06F16/24578G06F16/25G06F16/285G06F17/10

Abstract

Claims (54)

1 . A parallel database management system comprises:

a plurality of node clusters, wherein a node cluster of the plurality of node clusters s includes a set of nodes, wherein a node of the set of nodes includes a set of silo units, wherein a silo unit of the set of silo units includes:

a set of processing units;

memory operably coupled to the set of processing units;

a set of disk drives operably coupled to the set of processing units; and

a network interface operably coupled to the set of processing units;

wherein the silo unit stores, in memory and/or in the set of disk drives an operating system and a database management software application;

wherein, when silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters execute, substantially in parallel, the database management software application, the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters are operable to:

bypass the operating system to store, substantially in parallel, a massive volume of data directly into respective memories and/or respective set of disk drives of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters; and

bypass the operating system to read, substantially in parallel, a second massive volume of data directly from the respective memories and/or the respective set of disk drives of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters.

2 . The parallel database management system of claim 1 further comprises:

the set of nodes including one or more nodes;

the set of silo units includes one or more silo units;

the set of processing units includes one or more processing units; and/or

the set of disk drives includes one or more disk drives.

3 . The parallel database management system of claim 1 , wherein a disk drive of the set of disk drives comprises one of:

Non-volatile Random-Access Memory (NVRAM);

Serial Advanced Technology Attachment (SATA);

Solid State Drives (SSDs); and

Non-volatile Memory Express (NVMe).

4 . The parallel database management system of claim 1 further comprises:

the massive volume of data includes at least a portion of a dataset, wherein the dataset includes a plurality of records, and wherein a record of the plurality of records includes a plurality of columns of data; and

the second massive volume of data includes at least a second portion of the dataset, wherein the at least the second portion of the dataset includes at least some of the columns of data of at least some of the plurality of records.

5 . The parallel database management system of claim 4 further comprises:

the dataset is segmented into a plurality of data segments, wherein the plurality of data segments forms a plurality of groups of data segments, wherein a group of data segments of the plurality of groups of data includes a set of data segments, and wherein a data segment of the plurality of data segments corresponds to a set of records of the plurality of records;

wherein the plurality of node clusters stores the plurality of data segments;

wherein a first node cluster of the plurality of node clusters stores a first grouping of groups of data segments of the plurality of data segments;

wherein a first set of silo units of the first node cluster stores a first group of data segments of the first grouping of groups of data segments; and

wherein a first silo unit of the first set of silo units stores a first data segment of the first group of data segments.

6 . The parallel database management system of claim 1 further comprises:

the massive volume of data includes at least a portion of a dataset;

wherein the dataset includes a plurality of records;

wherein a record of the plurality of records includes a plurality of columns of data;

wherein the dataset is segmented into a plurality of data segments;

wherein the plurality of data segments forms a plurality of groups of data segments;

wherein a group of data segments of the plurality of groups of data includes a set of data segments;

wherein a data segment of the plurality of data segments corresponds to a set of records of the plurality of records;

wherein one or more columns of data of the plurality of data columns of data of the set of records of the data segment form a coding block; and

wherein coding blocks form a coding line, wherein the coding line is stored in the silo unit.

7 . The parallel database management system of claim 6 further comprises:

wherein, when silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters execute, substantially in parallel, the database management software application, the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters are operable to:

generate a plurality of manifests that correspond to the plurality of data segments, wherein a first manifest of the plurality of manifests corresponds to a first data segment of the plurality of data segments, and wherein the first manifest contains data regarding storage of the first data segment by a first silo unit of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters; and

store the plurality of manifests with the plurality of data segments.

8 . The parallel database management system of claim 7 , wherein the data of the first manifest comprises one or more of:

a physical location within the first silo unit as to where the first data segment is stored; and

a cluster key that is used to organize the first data segment based on a column associated with the cluster key.

9 . The parallel database management system of claim 6 further comprises:

wherein, when silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters execute, substantially in parallel, the database management software application, the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters are, prior to storing the massive volume of data, operable to:

allocate respective huge pages of virtual memory that is tied to physical memory for the direct data storage of the massive volume of data into the respective memories and/or the respective set of disk drives of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters.

10 . The parallel database management system of claim 1 further comprises:

wherein, when silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters execute, substantially in parallel, the database management software application, the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters are operable to:

query the operating system to identify the respective memories and/or the respective set of disk drives of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters;

query the operating system to identify respective sets of processing unit of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters; and

query the operating system to identify respective network interfaces of the silo-units-of-the-sets-of-silo-units-of-the-sets-of-nodes-of-the-node-clusters-of-the-plurality-of-node-clusters.

Assignments (2)

Continuity (5)

References Cited (84)