VM Reachability Convergence for Faster Data Center Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data center networks, particularly VXLAN EVPN fabrics, the convergence time for updating IP and MAC address locality of virtual machines (VMs) during relocation is elongated, leading to increased downtime and scalability challenges due to the need for per-VM GARP/ND messages and large numbers of top-of-rack switches, resulting in inefficient network connectivity.
Innovation Solution
Implementing techniques that include proactive convergence of endpoint reachability by determining VM relocation, storing and advertising MAC and IP addresses, utilizing a route reflector to update the network efficiently, and poisoning routes to the old location, thereby reducing convergence time and downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If per-VM GARP/ND messages are used for VM relocation, then endpoint reachability is updated, but convergence time increases
Solution Approach 1:
The patent applies preliminary action by proactively updating endpoint reachability information before VM relocation actually occurs. The system receives notifications of scheduled VM moves from the orchestrator and pre-updates the location bindings in the network fabric, so that when the VM moves, the network is already converged and ready to route traffic to the new location immediately.
Solution Approach 2:
The patent introduces an intermediary mechanism - a notification system from the orchestrator to the network fabric - that mediates between VM relocation events and network convergence. This intermediary allows the network to be informed of upcoming moves and proactively update routing information without waiting for traditional GARP/ND messages triggered by the VM itself.
2Adaptability or versatility
If multiple vNICs per VM are supported, then network functionality is improved, but convergence complexity increases
Solution Approach 1:
The patent merges the convergence process for multiple vNICs by receiving a single notification from the orchestrator that contains information about all vNICs associated with a VM. Instead of processing each vNIC separately with individual GARP/ND messages, the system handles them collectively through one proactive update mechanism, reducing complexity while maintaining support for multiple network interfaces.
3Productivity
If large numbers of ToRs are deployed, then network scalability is improved, but message propagation overhead increases
Solution Approach 1:
The patent uses the orchestrator notification as an intermediary that consolidates relocation information before it reaches the network fabric. This single notification event triggers coordinated updates across multiple ToRs, avoiding the need for each ToR to independently process and propagate separate GARP/ND messages for every VM move, thereby reducing overall message overhead in large-scale deployments.
Data Source
AI summary
This disclosure describes techniques and mechanisms to enable proactive convergence of endpoints in a network corresponding to virtual machine mobility scenarios. The techniques may be used individually or together to reduce convergence times and improve network security. The techniques may include proactively exchanging IP-MAC binding information between an old location and a new location of a virtual machine; (2) utilizing a route reflector and a convergence efficient flood order to update the location of the virtual machine at the old location to point to the new location; (3) proactively poisoning the route to the virtual machine at the old location; and (4) utilizing an orchestrator to proactively notify leaf(s) of reachability information of the virtual machine prior to or during migration of the virtual machine.