US 2017/0324812 A1Published application

Multi-layer data storage system

AssigneeMasergy Communications, Inc.

InventorMichael Roy Stute

Life of this patent

Estimating the term…

Loading patent…

Abstract

The disclosed embodiments include data storage systems and methods to store data. In one embodiment, the system includes a storage medium having a plurality of partitions for storing data values of a dataset. The system also includes a system manager operable to receive a data value connected to the system manager and assign a plurality of data subsets of the dataset to the plurality of partitions. The system also includes a key manager operable to assign a key of a key space to each data value of the dataset. The system also includes a top layer data manager and at least one bottom layer data manager, each bottom layer data manager assigned to a data subset. The top layer data manager passes data values to the at least one bottom layer data manager. The system also includes an allocation manager operable to dynamically readjust a storage space of the partition.

Claims32

Dependent claims

A multi-layer data storage system, comprising:

a storage medium having a plurality of partitions for storing data values of a dataset;
a system manager operable to:
receive, over a network, a data value of the dataset from an external source communicatively connected to the system manager, the dataset having a plurality of data subsets;
and
assign the plurality of data subsets of the dataset to the plurality of partitions;
a key manager operable to assign a key of a key space to each data value of the dataset, each data subset of the dataset being assigned at least one key of the key space;
a first top layer data manager operable to:
obtain a first set of keys of the key space;
and
divide the first set of keys into a first plurality of subsets of keys;
at least one bottom layer data manager, each bottom layer data manager of the at least one bottom layer data manager being assigned a data subset of the plurality of data subsets, and each bottom layer data manager operable to:
obtain at least one key of the first set of keys assigned to the data subset;
store the at least one key on a partition the data subset is assigned to;
provide an indication of a probability the data value belongs to the data subset based on whether a key of the first set of keys assigned to the data value is stored on the partition;
obtain the data value of the data subset;
and
store the obtained data value on the partition;
and
an allocation manager operable to:
determine whether to readjust a storage space of at least one partition of the plurality of partitions;
and
dynamically readjust the at least one partition based on at least one of a frequency data values of the dataset are stored on the at least one partition and a storage capacity of the at least one partition.

19 more elements

Dependencies are inferred from the claim text.