US 12,192,282 B1Granted

Method and system for multi-path network operations between storage clusters

AssigneeDataDirect Networks, Inc.

InventorsDouglas Oucharek, Silvain Didelot

Life of this patent

Estimating the term…

Loading patent…

Abstract

Multi-Path approach to data migration between nodes in distributed data storage systems is rendered by creating multiple connections between application cores across interfaces on both an initiator node and a target node and load balancing whole messages and RDMA operations over the various pathways routed via the same or different network fabric types. The Multi-Path approach adapts to applications which are core-affine and those which are not. The balancing of whole messages/RDMA operations is performed by a send/receive API command specifying the target node for the whole message (and optionally a core number in the case of a core-affine target). The Multi-Path approach adapts dynamically to network errors by excluding faulty pathways, as needed, and shifting traffic to remaining healthy pathways with background recovery of pathways running periodically for self-healing. The Multi-Path also adapts dynamically to the load by diverting traffic from slow pathways and favoring pathways with normal load.

Claims20

Dependent claims

A method for migrating data between compute nodes in a distributed data storage system, comprising:

configuring a distributed storage system with a plurality of compute nodes and a Multi-Path sub-system integrated in said distributed storage system in operative coupling with said plurality of compute nodes,
wherein said plurality of compute nodes include at least one initiating node and at least one target node, wherein said Multi-Path sub-system is configured with a plurality of initiating interface units incorporated in said at least one initiating node and a plurality of target interface units incorporated in said at least one target node, and
wherein a plurality of communication pathways are operatively coupled between said plurality of initiating interface units and said plurality of target interface units;
forming, by said distributed storage system, at least one whole message and incorporating an identification of said at least one target node in said at least one whole message, wherein said at least one whole message is based on a Remote Procedure Call (RPC) protocol;
selecting, by said Multi-Path sub-system, at least one first pathway from said plurality of communication pathways for sending said at least one whole message;
sending said at least one whole message from said at least one initiating node to said at least one target node over said at least one first pathway;
selecting, by said Multi-Path sub-system, at least one second pathway from remaining communication pathways of said plurality of communication pathways for sending said at least one whole message if the communication over said at least one first pathway has failed, wherein said at least one first and at least one second pathways are selected from a group consisting of the pathways routed via the same network fabric type, different network fabric types, and a combination thereof;
and
sending said at least one whole message from said at least one initiating node to said at least one target node over said at least one second pathway.

6 more elements

A distributed data storage system for migrating data between compute nodes, comprising:

a plurality of compute nodes and a Multi-Path sub-system integrated in said distributed data storage system in operative coupling with said plurality of compute nodes, wherein said plurality of compute nodes include at least one initiating node and at least one target node, where said at least one initiating node comprises a first memory and a first processor and said at least one target node comprises a second memory and a second processor,
wherein said Multi-Path sub-system is configured with a plurality of initiating interface units incorporated in said at least one initiating node and a plurality of target interface units incorporated in said at least one target node, and
wherein a plurality of communication pathways are operatively coupled between said plurality of initiating interface units and said plurality of target interface units;
said distributed storage system being configured for migrating at least one whole message incorporating an identification of said at least one target node from said at least one initiating node to said at least one target node over at least one first pathway of said plurality of communication pathways and at least one second pathway of said plurality of communication pathways if said at least one first pathway has failed, wherein said at least one first and at least one second pathways are selected from a group consisting of the pathways routed via the same network fabric type, different network fabric types, and a combination thereof.

1 more element

Dependencies are inferred from the claim text.

Facts

Application: 18/504,434Filed 8 Nov 2023
Granted: 7 Jan 2025
Term ends: Not estimatedThe source records don’t support an estimate.

Classification5 codes

H04L 67/1097
In which an application is distributed across nodes in the network: for distributed storage of data in networks, e.g. transport arrangements for…
H04L 45/28
Routing or path finding of packets in data switching networks: using route fault recovery
H04L 67/1038
For accessing one among a plurality of replicated servers: load balancing arrangements to avoid a single path through a load balancer
H04L 67/133
Protocols: protocols for remote procedure calls [RPC]

Family

No related US applications are recorded in the indexed filing history.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key