Flow Rule Virtualization for Programmable Network Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Programmable network virtualization (P-NV) faces scalability issues due to high resource consumption and control channel usage, primarily because existing network hypervisors install flow rules in a one-to-one manner, leading to increased CPU cycles and memory usage.
Innovation Solution
A flow rule virtualization method that generates a virtualization type table to select a virtualization type for mapping virtual flow rules to physical flow rules, merges virtual flow rules into a smaller number of physical rules using a rule map, and installs the virtualized physical flow rules on physical switches, reducing resource consumption and optimizing network management policies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If flow rules are installed in a one-to-one manner from virtual network to physical network, then network management flexibility and programmability are improved, but resource consumption (CPU cycles, control channel bandwidth, memory) increases significantly
Solution Approach 1:
The patent merges multiple virtual flow rules into a single physical flow rule by identifying common characteristics among virtual flow rules (such as same match fields and actions) and consolidating them. This reduces the total number of flow rules that need to be installed and managed on physical switches, thereby decreasing CPU cycles, control channel bandwidth consumption, and memory usage while preserving the ability to manage virtual networks flexibly through the virtualization layer
2Adaptability or versatility
If the number of flow rules is increased to support more virtual network operations, then network functionality and tenant customization are improved, but control channel bandwidth and CPU usage increase
Solution Approach 1:
The patent combines multiple virtual flow rules that share common match fields and actions into consolidated physical flow rules. This merging process reduces the quantity of flow rules transmitted over the control channel, thereby decreasing control channel bandwidth consumption while maintaining full network functionality through the virtualization abstraction layer that manages the mapping between virtual and physical rules
3Quantity of substance
If compression techniques are used to decrease memory used to store flow rules, then memory consumption is reduced, but delay increases and resource consumption increases during setup
Solution Approach 1:
The patent merges virtual flow rules into fewer physical flow rules at the virtualization layer, which inherently reduces the memory required to store flow rules without requiring additional compression operations. This approach avoids the time delay and setup resource consumption associated with compression techniques, as the reduction is achieved through logical consolidation rather than data compression
4Quantity of substance
If blind compression of flow rules is applied, then memory usage decreases, but tenant semantics may be violated and network management policies are compromised
Solution Approach 1:
The patent merges virtual flow rules into physical flow rules through a controlled virtualization process that preserves tenant semantics and network management policies. The merging is performed by the network hypervisor with full knowledge of the virtual network configuration and tenant requirements, ensuring that consolidation does not violate any semantic constraints or policy requirements, unlike blind compression approaches
Data Source
AI summary
A flow rule virtualization method is disclosed. The flow rule virtualization method may include: (a) generating a virtualization type table, (b) selecting a virtualization type, which represents the degree of mapping between a virtual flow rule and a physical flow rule executed on a physical network, for a virtual flow rule generated from a tenant controller by using the virtualization type table, (c) virtualizing the virtual flow rule into the physical flow rule according to the selected virtualization type, (d) virtualizing the physical rule by using a rule map, and (e) installing the virtualized physical flow rule on the physical switch.