Granted Patent B2

US 11,895,093 · App. 16/936,559 · Granted Feb 6, 2024

Inventors: Dharmesh M. Patel (Round Rock, TX); Ravikanth Chaganti (Bangalore, IN)

Assignee: Dell Products L.P.

H04L63/029G06F9/5083H04L12/66H04L67/10H04L67/1097H04L67/563H04L67/60

Abstract

Claims (77)

1. A method for accessing data nodes of a data cluster, the method comprising:

initiating, by a data processor, performance of a data access gateway (DAG) on a first data node of data nodes upon obtaining system metadata, wherein:

the system metadata comprises prior performance information specifying at least one latency and at least one throughput associated with previous DAG operations performed by at least one of the data nodes; and

the previous DAG operations performed by the at least one of the data nodes comprise:

obtaining a previous request of requests types from a host;

in response to obtaining the previous request, obtaining previous discovery metadata from each of the data nodes;

updating, upon obtaining the previous discovery metadata, a previous discovery metadata repository on the at least one of the data nodes; and

identifying, according to the previous discovery metadata, other data nodes of the data nodes to send the previous request;

in response to initiating the performance of the DAG on the first data node:

obtaining, by the DAG, a request of the request types from the host, wherein the DAG is not executing on the host, wherein the request types comprise:

a read request, a write request, an update request, an information request, and a delete request;

in response to obtaining the request:

obtaining discovery metadata from each data node of the data nodes in the data cluster;

wherein the discovery metadata specifies the data nodes of the data nodes that are available to serve the request;

updating, using the obtained discovery metadata, a discovery metadata repository;

identifying a second data node, of the data nodes, to serve the request according to the updated discovery metadata; and

sending the request to the second data node of the data nodes.

2. The method of claim 1 , the method further comprising:

following sending the request to the second data node:

making a first determination that the second data node is unable to serve the request;

in response to the first determination, retrying the request with the second data node and updating a retry count in the discovery metadata repository;

making a second determination that the second data node is unable to serve the request;

in response to the second determination, making a third determination that a retry limit associated with the second data node has been exceeded; and

in response to the third determination, sending the request to a third data node based on the discovery metadata.

3. The method of claim 2 , wherein the retry limit indicates a maximum allowable number of request retries for the second data node.

4. The method of claim 1 , wherein the system metadata further specifies computational resources available on the data nodes.

5. A system for accessing data nodes of a data cluster, comprising:

A physical processor configured to initiate performance of a data access gateway (DAG) on a first data node of data nodes upon obtaining system metadata, wherein:

the system metadata comprises prior performance information specifying at least one latency and at least one throughput associated with previous DAG operations performed by at least one of the data nodes; and

the previous DAG operations performed by the at least one of the data nodes comprise:

obtaining a previous request of requests types from a host;

in response to obtaining the previous request, obtaining previous discovery metadata from each of the data nodes;

updating, upon obtaining the previous discovery metadata, a previous discovery metadata repository on the at least one of the data nodes; and

identifying, according to the previous discovery metadata, other data nodes of the data nodes to send the previous request;

a data access gateway (DAG), which when executed by the processor performs a method, the method comprising:

obtaining, by the DAG, a request of the request types from the host, wherein the DAG is not executing on the host, wherein the request types comprise:

a read request, a write request, an update request, an information request, and a delete request;

in response to obtaining the request:

obtaining discovery metadata from each data node of the data nodes in the data cluster;

wherein the discovery metadata specifies the data nodes of the data nodes that are available to serve the request;

updating, using the obtained discovery metadata, a discovery metadata repository;

identifying a second data node, of the data nodes, to serve the request according to the updated discovery metadata; and

sending the request to the second data node of the data nodes.

6. The system of claim 5 , wherein the method further comprising:

following sending the request to the second data node:

making a first determination that the second data node is unable to serve the request;

in response to the first determination, retrying the request with the second data node and updating a retry count in the discovery metadata repository;

making a second determination that the second data node is unable to serve the request;

in response to the second determination, making a third determination that a retry limit associated with the second data node has been exceeded; and

in response to the third determination, sending the request to a third data node based on the discovery metadata.

7. The system of claim 6 , wherein the retry limit indicates a maximum allowable number of request retries for the second data node.

8. The system of claim 5 , wherein the system metadata further specifies computational resources available on the data nodes of the data nodes.

9. A non-transitory computer readable medium comprising computer readable computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for accessing data nodes of a data cluster, the method comprising:

initiating, by a data processor, performance of a data access gateway (DAG) on a first data node of data nodes upon obtaining system metadata, wherein:

the system metadata comprises prior performance information specifying at least one latency and at least one throughput associated with previous DAG operations performed by at least one of the data nodes; and

the previous DAG operations performed by the at least one of the data nodes comprise:

obtaining a previous request of requests types from a host;

in response to obtaining the previous request, obtaining previous discovery metadata from each of the data nodes;

updating, upon obtaining the previous discovery metadata, a previous discovery metadata repository on the at least one of the data nodes; and

identifying, according to the previous discovery metadata, other data nodes of the data nodes to send the previous request;

in response to initiating the performance of the DAG on the first data node:

obtaining, by the DAG, a request of the request types from the host, wherein the DAG is not executing on the host, wherein the request types comprise:

a read request, a write request, an update request, an information request, and a delete request;

in response to obtaining the request:

obtaining discovery metadata from each data node of the data nodes in the data cluster;

wherein the discovery metadata specifies the data nodes of the data nodes that are available to serve the request;

updating, using the obtained discovery metadata, a discovery metadata repository;

identifying a second data node, of the data nodes, to serve the request according to the updated discovery metadata; and

sending the request to the second data node of the data nodes.

10. The non-transitory computer readable medium of claim 9 , wherein the method further comprising:

following sending the request to the first data node:

making a first determination that the second data node is unable to serve the request;

in response to the first determination, retrying the request with the second data node and updating a retry count in the discovery metadata repository;

making a second determination that the second data node is unable to serve the request;

in response to the second determination, making a third determination that a retry limit associated with the second data node has been exceeded; and

in response to the third determination, sending the request to a third data node based on the discovery metadata.

11. The non-transitory computer readable medium of claim 10 , wherein the retry limit indicates a maximum allowable number of request retries for the second data node.

Assignments (9)