Virtual Distributed Bridging for Hybrid Network Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network technologies face challenges in seamlessly bridging between logical networks (such as VXLAN) and physical networks (such as VLAN), particularly in hybrid environments where L2 centric protocols are needed, and there is a lack of efficient mechanisms for transitioning workloads between these networks.
Innovation Solution
The implementation of a virtual distributed router/bridge (VDRB) module that performs L3 routing and bridging operations, with virtual distributed bridges (VDBs) partitioning bridging tasks among host machines based on MAC addresses, using sharding to distribute MAC addresses and dynamically reassign them in case of failures or changes, and employing mechanisms to prevent packet loops and duplicates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If bridging tasks are centralized in a single network controller, then management and control are simplified, but the system lacks scalability and becomes a single point of failure
Solution Approach 1:
The patent segments the centralized bridging function into distributed Virtual Distributed Bridges (VDBs) across multiple host machines. Each VDB handles a specific shard of MAC addresses, transforming the single-point control model into a distributed architecture that maintains manageability while achieving scalability and fault tolerance through replication across hosts.
Solution Approach 2:
The patent introduces a new dimension of distribution by deploying VDB instances across multiple spatial locations (host machines) rather than concentrating control in a single controller. This dimensional shift from centralized to distributed architecture enables the system to scale horizontally while maintaining the organizational simplicity of centralized management through the controller's role in coordinating VDBs.
2Reliability
If all VDB instances bridge all packets, then comprehensive bridging coverage is achieved, but network performance degrades due to redundant processing and potential loops
Solution Approach 1:
The patent segments the MAC address space into distinct shards, with each VDB instance responsible for bridging only packets containing MAC addresses within its assigned shard. This segmentation eliminates redundant bridging operations across VDB instances while maintaining complete bridging coverage, as every MAC address belongs to exactly one shard and is handled by the appropriate VDB.
Solution Approach 2:
Each VDB instance is assigned specific local responsibility for a particular shard of MAC addresses, allowing it to optimize its bridging operations for that specific subset. This local quality approach ensures that each VDB processes only relevant packets, improving overall network throughput while maintaining reliable bridging coverage through the collective capability of all VDB instances.
3Device complexity
If MAC address shards are statically assigned to VDBs, then system simplicity is maintained, but the system cannot adapt to failures or dynamic workload changes
Solution Approach 1:
The patent implements dynamic shard assignment where the network controller can reassign MAC address shards to different VDB instances based on current system conditions. When a VDB fails or workload changes, the controller dynamically redistributes shards to maintain optimal performance and availability, transforming the static assignment model into an adaptive system that responds to runtime conditions.
Solution Approach 2:
The network controller monitors the health and performance of VDB instances and uses this feedback to make informed decisions about shard assignment. When failures or workload changes are detected, the controller adjusts shard distribution accordingly, creating a closed-loop system that continuously adapts to maintain optimal bridging performance and availability.
4Reliability
If distributed VDB architecture is implemented, then scalability and fault tolerance are improved, but system complexity increases due to coordination and sharding management
Solution Approach 1:
The network controller serves as an intermediary that manages the complexity of the distributed VDB system. It handles shard assignment, monitors VDB health, and coordinates failover operations, thereby shielding individual VDB instances from the complexity of distributed system management while enabling them to benefit from the fault tolerance and scalability of the distributed architecture.
Data Source
AI summary
Virtualization software that includes a VDRB (virtual distributed router/bridge) module for performing L3 routing and/or bridging operations is provided. At least some of the VDRBs are configured as VDBs (virtual distributed bridge) for performing bridging operations between different network segments in a distributed manner. The bridging tasks of a network are partitioned among several VDBs of the network based on MAC addresses. MAC addresses of VMs or other types of network nodes belonging to an overlay logical network are partitioned into several shards, each shard of MAC addresses assigned to a VDB in the network. Each VDB assigned a shard of MAC addresses performs bridging when it receives a packet bearing a MAC address belonging to its assigned shard. A VDB does not perform bridging on packets that do not have MAC address that falls within the VDB's shard of MAC addresses.