Granted Patent B2

US 11,379,118 · App. 17/081,743 · Granted Jul 5, 2022

Inventors: Kedar Sadanand Godbole (Pune, IN); Nitin Madan (Cupertino, CA); Srikant Viswanathan (Pune, IN)

Assignee: EMC IP Holding Company LLC

G06F3/0604G06F3/067G06F3/0652

Abstract

Claims (67)

1. A method for storage load balancing between storage nodes, comprising:

monitoring a storage utilization on a source storage node;

making a first determination that the storage utilization at least meets a warning threshold;

generating, based on the first determination and from virtual synthetics metadata, a relationship graph for files maintained on the source storage node;

identifying a first subset of the files based on the relationship graph; and

migrating the first subset of the files to a destination storage node.

2. The method of claim 1 , wherein the virtual synthetics metadata for a file comprises at least one metadata record, wherein each metadata record of the at least one metadata record comprises information mapping a first file segment of the file to a second file segment of a base file, wherein the first and second file segments comprise identical content.

3. The method of claim 2 , wherein the information comprises a base file identifier associated with the base file, a base file segment offset for the second file segment, a base file segment length for the second file segment, and a file offset for the first file segment.

4. The method of claim 1 , wherein the relationship graph is a disconnected graph comprising at least two independent sub-graphs, wherein each sub-graph is a connected graph.

5. The method of claim 1 , wherein generating the relationship graph for files maintained on the source storage node from the virtual synthetics metadata, comprises:

for each file associated with the virtual synthetics metadata, of the files maintained on the source storage node:

adding, if not existing already, a first vertex representative of the file to the relationship graph;

examining the virtual synthetics metadata associated with the file to identify at least one base file;

adding, for each base file of the at least one base file and if not existing already, a second vertex representative of the base file to the relationship graph; and

adding, for each base file of the at least one base file, an edge connecting the first vertex to the second vertex.

6. The method of claim 1 , wherein identifying the first subset of the files based on the relationship graph, comprises:

selecting a relationship sub-graph of the relationship graph; and

identifying, for each vertex forming the relationship sub-graph, a file being represented by the vertex.

7. The method of claim 1 , further comprising:

obtaining an updated storage utilization for the source storage node;

making a second determination that the updated storage utilization exceeds a goal threshold;

identifying, based on the second determination, a second subset of the files based on the relationship graph; and

migrating the second subset of the files to the destination storage node.

8. A non-transitory computer readable medium (CRM) comprising computer readable program code, which when executed by a computer processor, enables the computer processor to:

monitor a storage utilization on a source storage node;

make a first determination that the storage utilization at least meets a warning threshold;

generate, based on the first determination and from virtual synthetics metadata, a relationship graph for files maintained on the source storage node;

identify a first subset of the files based on the relationship graph; and

migrate the first subset of the files to a destination storage node.

9. The non-transitory CRM of claim 8 , wherein the virtual synthetics metadata for a file comprises at least one metadata record, wherein each metadata record of the at least one metadata record comprises information mapping a first file segment of the file to a second file segment of a base file, wherein the first and second file segments comprise identical content.

10. The non-transitory CRM of claim 9 , wherein the information comprises a base file identifier associated with the base file, a base file segment offset for the second file segment, a base file segment length for the second file segment, and a file offset for the first file segment.

11. The non-transitory CRM of claim 8 , wherein the relationship graph is a disconnected graph comprising at least two independent sub-graphs, wherein each sub-graph is a connected graph.

12. The non-transitory CRM of claim 8 , further comprising computer readable program code to generate the relationship graph, which when executed by the computer processor, further enables the computer processor to:

for each file associated with the virtual synthetics metadata, of the files maintained on the source storage node:

add, if not existing already, a first vertex representative of the file to the relationship graph;

examine the virtual synthetics metadata associated with the file to identify at least one base file;

add, for each base file of the at least one base file and if not existing already, a second vertex representative of the base file to the relationship graph; and

add, for each base file of the at least one base file, an edge connecting the first vertex to the second vertex.

13. The non-transitory CRM of claim 8 , further comprising computer readable program code to identify the first subset of the files, which when executed by the computer processor, further enables the computer processor to:

select a relationship sub-graph of the relationship graph; and

identify, for each vertex forming the relationship sub-graph, a file being represented by the vertex.

14. The non-transitory CRM of claim 8 , further comprising computer readable program code, which when executed by the computer processor, further enables the computer processor to:

obtain an updated storage utilization for the source storage node;

make a second determination that the updated storage utilization exceeds a goal threshold;

identify, based on the second determination, a second subset of the files based on the relationship graph; and

migrate the second subset of the files to the destination storage node.

15. A system, comprising:

a plurality of storage nodes, comprising:

a destination storage node; and

a source storage node operatively connected to the destination storage node, and comprising a computer processor,

wherein a data migration agent executes on the computer processor and is programmed to:

monitor a storage utilization on the source storage node;

make a first determination that the storage utilization at least meets a warning threshold;

generate, based on the first determination and from virtual synthetics metadata, a relationship graph for files maintained on the source storage node;

identify a first subset of the files based on the relationship graph; and

migrate the first subset of the files to the destination storage node.

16. The system of claim 15 , wherein the virtual synthetics metadata for a file comprises at least one metadata record, wherein each metadata record of the at least one metadata record comprises information mapping a first file segment of the file to a second file segment of a base file, wherein the first and second file segments comprise identical content.

17. The system of claim 16 , wherein the information comprises a base file identifier associated with the base file, a base file segment offset for the second file segment, a base file segment length for the second file segment, and a file offset for the first file segment.

18. The system of claim 15 , wherein the relationship graph is a disconnected graph comprising at least two independent sub-graphs, wherein each sub-graph is a connected graph.

19. The system of claim 15 , wherein to identify the first subset of the files, the data migration agent is further programmed to:

select a relationship sub-graph of the relationship graph; and

identify, for each vertex forming the relationship sub-graph, a file being represented by the vertex.

20. The system of claim 15 , wherein the data migration agent is further programmed to:

obtain an updated storage utilization for the source storage node;

make a second determination that the updated storage utilization exceeds a goal threshold;

identify, based on the second determination, a second subset of the files based on the relationship graph; and

migrate the second subset of the files to the destination storage node.

Assignments (9)

Continuity (1)