Live Migration Packet Duplication for VM Network Continuity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtual computing environments, live migration of virtual machines often results in packet loss due to delayed propagation of updated network mappings, causing disruptions as encapsulation protocol entities may reject packets if they do not have the latest mapping, especially in large-scale networks with tens of thousands of hosts and devices.
Innovation Solution
A packet duplication technique is employed during live migration, where the migration-destination virtualization host generates duplicate encapsulation packets for outbound packets, ensuring that packets are delivered even if the updated mapping has not been propagated to all entities, by making one packet appear to originate from the original host and another from the new host, thus maintaining continuous communication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If live migration is performed to enable load balancing and maintenance operations, then system flexibility and maintenance capability are improved, but packet loss and communication disruptions occur due to delayed mapping propagation
Solution Approach 1:
The system performs preliminary actions by setting up packet duplication mechanisms before the mapping propagation completes. The migration-destination host prepares to send duplicate packets anticipating that some entities may still have old mappings, preventing packet loss before it occurs.
Solution Approach 2:
The system creates copies of network packets from the migration-destination host that mimic packets from the migration-source host. These duplicate packets carry the same payload but have different source identifiers, allowing recipients with outdated mappings to accept packets from either source.
2Manufacturing precision
If updated network mappings are propagated quickly to all entities, then packet delivery accuracy is improved, but the complexity of coordination among numerous hosts and devices increases
Solution Approach 1:
The system enables self-service by allowing the migration-destination host to autonomously generate and send duplicate packets without requiring complex coordination with every entity in the network. Each entity independently determines whether to accept packets based on its local mapping state, eliminating the need for centralized coordination.
Solution Approach 2:
The system segments the packet delivery problem into two independent paths: packets from the migration-source host and duplicate packets from the migration-destination host. This segmentation allows each path to operate independently based on local mapping state, avoiding the need for global coordination.
3Reliability
If packet duplication is implemented during live migration, then packet loss is minimized and communication continuity is maintained, but network traffic volume and processing overhead increase
Solution Approach 1:
The system applies partial action by sending duplicate packets only from the migration-destination host, not from all hosts. This selective duplication provides sufficient redundancy to handle mapping propagation delays without the excessive overhead of universal packet replication.
Solution Approach 2:
The system implements discarding and recovering by allowing duplicate packets to be sent and then discarded if the recipient already has the original packet. Recipients can recover from potential packet loss by accepting either the original or duplicate packet, eliminating the need for retransmission protocols.
Data Source
AI summary
A virtualization management component of a particular host at a virtualized computing determines that a packet duplication phase of a migration of a virtual machine from a source host to the particular host has begun. The virtualization management component detects that a migrated version of the virtual machine, instantiated at the particular virtualization host, has generated a baseline packet directed to a destination address. The virtualization management component sends a first encapsulation packet comprising the baseline packet to an encapsulation intermediary associated with the destination address. The virtualization management component sends a second encapsulation packet comprising the baseline packet to the source virtualization host, with a directive to forward the second encapsulation packet to the encapsulation intermediary.