Software-Defined Network Scaling via Centralized Policy Redistribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional load balancing methods in data networks are inefficient due to static devices managing packet distribution, leading to wasted resources and inability to dynamically scale up or down, resulting in suboptimal resource utilization.
Innovation Solution
A method and system for scaling data networks by generating and sending service policies across nodes to reassign service requests, including the creation of redirection policies to manage heavy-duty connections, allowing for seamless addition or removal of nodes without interrupting existing connections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional static load balancers are used to manage packet distribution, then device simplicity is maintained, but network adaptability and resource utilization deteriorate due to inability to dynamically scale
Solution Approach 1:
The system segments the load balancing functionality by separating the control plane (centralized controller) from the data plane (switches and servers). The centralized controller manages policy distribution and coordination, while switches execute local forwarding decisions based on received policies. This segmentation enables dynamic scaling capabilities without requiring complex modifications to individual switching devices.
Solution Approach 2:
A centralized controller acts as an intermediary between the management system and network devices. The controller receives scaling requests, computes appropriate service policies, and distributes them to relevant switches and servers. This intermediary approach abstracts the complexity of dynamic scaling from individual devices while maintaining overall system adaptability.
2Productivity
If static load balancers are used, then device complexity is reduced, but resource utilization deteriorates due to wasted capacity during low load periods
Solution Approach 1:
The system implements dynamic load balancing policies that automatically adjust traffic distribution based on current network conditions and resource availability. When scaling events occur (nodes added or removed), the centralized controller computes updated service policies and distributes them to switches, enabling the network to dynamically optimize resource utilization without manual intervention.
Solution Approach 2:
The system incorporates feedback mechanisms where switches report traffic statistics and node status to the centralized controller. The controller uses this feedback information to compute optimized service policies that improve resource utilization. This closed-loop feedback enables the network to adapt to changing conditions and eliminate wasted capacity.
3Productivity
If nodes are removed from the network, then resource allocation is optimized, but service continuity deteriorates due to interruption of existing connections
Solution Approach 1:
Before removing a node from the network, the centralized controller pre-computes and distributes updated service policies to relevant switches. These policies prepare the network for the node removal by establishing alternative traffic paths in advance. When the node is actually removed, service continuity is maintained because the switches are already configured to forward traffic through alternative paths.
Solution Approach 2:
The centralized controller serves as an intermediary that coordinates the node removal process. It manages the transition by computing redistribution policies that route traffic away from the departing node through alternative paths. This coordinated approach ensures that service continuity is maintained throughout the removal process while still achieving resource optimization.
4Adaptability or versatility
If dynamic scaling is implemented, then network adaptability improves, but system complexity increases due to policy management requirements
Solution Approach 1:
The system extracts the complex policy management functionality from individual network devices and centralizes it in the centralized controller. The controller handles all policy computation, coordination, and distribution tasks, while switches focus on executing forwarding decisions based on received policies. This extraction reduces device complexity while maintaining dynamic scaling capabilities.
Solution Approach 2:
The system implements self-service capabilities where the centralized controller automatically computes and distributes service policies in response to scaling events without manual intervention. The controller monitors network state changes, triggers policy recomputation when needed, and pushes updated policies to relevant devices. This automation reduces the operational complexity of managing dynamic scaling.
Data Source
AI summary
Provided are methods and systems for graceful scaling of data networks. In one example, an indication of removal of a node from a plurality of nodes of the data network is received. A service policy is generated to reassign service requests associated with the node to another node in the plurality of nodes. The service policy is then sent to each of the plurality of nodes of the data network. To scale out a data network, an indication of presence of a further node in the data network is received, and a further node service policy is generated and sent to each of the plurality of nodes of the data network and to the further node. Additional actions can be taken in order to prevent interruption of an existing heavy-duty connection while scaling the data network.