Reverse Replication for Live Migration Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing live migration technologies face challenges in maintaining consistency across virtual machines during migration, particularly when simultaneous freezing of VMs is not feasible, leading to potential data corruption and service interruptions.
Innovation Solution
The implementation of reverse replication, where writes to migrated virtual machines are intercepted and mirrored back to the source host, ensuring consistency through active/active asynchronous replication, and the use of virtual volumes and virtual storage array networks to manage disk access and ownership transfer without disk copying.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If live migration is performed without freezing VMs, then service continuity is maintained, but data consistency cannot be ensured
Solution Approach 1:
The system performs preliminary actions by freezing the VM on the source host before migration, and uses reverse replication to pre-establish the consistency group at the source host. This preliminary preparation ensures that when the VM is migrated, the source host already has a consistent baseline to replicate writes back to, resolving the contradiction between maintaining service continuity and ensuring data consistency.
Solution Approach 2:
The system implements feedback through reverse replication, where writes to the migrated VM are continuously monitored and replicated back to the source host's consistency group. This feedback mechanism ensures that the source host maintains an up-to-date copy of the VM's data, allowing the system to resolve data consistency issues that arise during live migration while preserving service continuity.
2Reliability
If VMs are frozen simultaneously during migration, then data consistency is maintained, but service interruption occurs
Solution Approach 1:
The system freezes only the source VM before migration rather than requiring simultaneous freezing of all VMs. The reverse replication mechanism is pre-configured to handle write operations, allowing the destination VM to accept writes immediately after migration while the source host maintains a consistency group. This preliminary setup eliminates service interruption while preserving data consistency.
Solution Approach 2:
The consistency group at the source host acts as an intermediary that receives replicated writes from the migrated VM. This intermediary structure allows the VM to be migrated without freezing, as the consistency group mediates between the destination host's write operations and the source host's data storage, maintaining consistency without service interruption.
3Reliability
If reverse replication is implemented, then source host consistency is maintained, but system complexity increases
Solution Approach 1:
The reverse replication mechanism serves multiple functions: it maintains the consistency group at the source host, enables failback capability, and provides a recovery point for migration failures. By making this single mechanism multi-functional, the system achieves source host consistency without proportionally increasing complexity, as the same infrastructure supports multiple reliability goals.
Solution Approach 2:
The system creates a copied consistency group at the source host that mirrors the migrated VM's data. This copying approach simplifies the complexity by using a straightforward replication mechanism rather than requiring complex coordination protocols. The copied consistency group provides the necessary reliability while keeping the system architecture relatively simple through the use of standard replication techniques.
4Reliability
If disk copying is performed during migration, then data consistency is ensured, but migration time increases
Solution Approach 1:
The system performs preliminary actions by setting up the reverse replication mechanism before the migration begins. The consistency group is pre-configured at the source host, and the replication infrastructure is established in advance. This preliminary preparation eliminates the need for time-consuming disk copying during the actual migration, as writes are handled through the pre-configured reverse replication path, thus reducing migration time while maintaining data consistency.
Solution Approach 2:
The reverse replication mechanism enables continuous handling of write operations throughout the migration process. Instead of stopping writes to perform copying, the system maintains continuous write handling by routing them through the consistency group. This continuity of useful action eliminates migration delays caused by data copying, as the replication process runs concurrently with normal VM operations, ensuring data consistency without increasing migration time.
Data Source
AI summary
Examples maintain consistency of writes for a plurality of VMs during live migration of the plurality from a source host to a destination host. The disclosure intercepts I/O writes to a migrated VM at a destination host and mirrors the I/O writes back to the source host. This “reverse replication” ensures that the CG of the source host is up to date, and that the source host is safe to fail back to if the migration fails.