Distributed Storage Control Nodes for Low-Latency Object Creation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage systems with multiple control nodes, increased client access leads to higher link bandwidth consumption for forwarding operation requests between control nodes, resulting in elevated access latency.
Innovation Solution
Implementing a method where a first control node receives and locally creates objects based on operation requests, with subsequent objects associated with the first, and transferring objects to other control nodes when load or capacity thresholds are exceeded to balance load and reduce bandwidth consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If operation requests are forwarded between control nodes, then the storage system can handle client access, but link bandwidth is consumed and access latency increases
Solution Approach 1:
The patent segments the storage system into multiple control nodes, each capable of independently handling client requests. By distributing request processing across multiple nodes rather than requiring all requests to be handled by a single node, the system reduces the need for inter-node forwarding and thereby decreases access latency while maintaining high client access capability.
Solution Approach 2:
The patent implements local quality by enabling each control node to independently create and manage objects locally. When a control node receives a client request, it can directly create objects on its local storage nodes without forwarding requests to other control nodes. This local processing approach eliminates inter-node communication overhead and reduces access latency.
2Quantity of substance
If multiple control nodes are used, then system capacity is increased, but link bandwidth occupation increases due to request forwarding
Solution Approach 1:
The patent implements self-service by enabling each control node to independently handle object creation and management operations. Control nodes do not need to forward requests to other nodes for these operations, as they can perform them locally. This self-service capability reduces inter-node communication and thereby decreases link bandwidth consumption while maintaining expanded system capacity through multiple nodes.
3Loss of time
If objects are created on local storage nodes, then access latency is reduced, but load balancing among control nodes may be affected
Solution Approach 1:
The patent applies dynamics by implementing a dynamic load balancing mechanism that adapts to the current system state. The load balancing algorithm considers factors such as the number of objects on each control node, storage capacity utilization, and current load levels. When creating objects, the system dynamically selects target control nodes based on real-time conditions, ensuring balanced distribution while allowing local object creation to proceed with minimal latency impact.
4Productivity
If control nodes handle requests locally, then access performance is optimized, but object distribution across the system may become uneven
Solution Approach 1:
The patent implements feedback through a load balancing mechanism that continuously monitors the distribution of objects across control nodes. When object creation requests are received, the system uses this feedback information to intelligently select target control nodes, preventing any single node from becoming overloaded. This feedback-driven approach maintains uniform object distribution while allowing control nodes to handle requests locally for optimal access performance.
Data Source
AI summary
Example methods and apparatus for managing a storage system are described. The storage system includes a plurality of control nodes. One example method includes receiving a first operation request of a client by a first control node in the plurality of control nodes. The first control node creates a first object on a first storage node based on the first operation request.