VM Snapshot Failover Sequencing for Lossless Data Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Failover procedures in data management systems often result in application downtime and data loss due to new data being written to the source environment between the time of snapshot capture and application migration to a target environment.
Innovation Solution
A failover procedure that involves capturing a snapshot of the application at the source, replicating it to a target environment, powering off the source, capturing an additional snapshot with minimal new data, and then powering on the application at the target, thereby reducing downtime and eliminating data loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a traditional failover procedure is used with single snapshot capture, then the backup process is simple, but application downtime increases and data loss occurs
Solution Approach 1:
The failover process is divided into multiple phases: capturing an initial snapshot before failover, capturing a second snapshot after powering off the source VM, transferring the first snapshot to the target environment, and applying incremental changes from the second snapshot. This segmentation allows the system to minimize downtime by capturing changes in discrete incremental steps rather than requiring a single large snapshot operation.
Solution Approach 2:
The system captures an initial snapshot of the virtual machine before initiating the failover process. This preliminary action ensures that the target environment has a baseline copy of the VM state, allowing the failover to proceed with minimal interruption. The preliminary snapshot is then used as the foundation for subsequent incremental updates during the failover process.
2Reliability
If snapshot capture is performed before failover, then data loss is prevented, but replication time increases
Solution Approach 1:
The snapshot process is segmented into two distinct phases: an initial snapshot captured before failover containing the complete VM state, and a second snapshot captured after powering off containing only incremental changes. By segmenting the replication process this way, the system can transfer the bulk of the data in the initial snapshot while only replicating minimal incremental changes in the second snapshot, significantly improving overall replication speed.
Solution Approach 2:
The system captures a second snapshot that contains only the incremental changes made to the virtual machine after the initial snapshot but before failover. This partial action approach focuses replication resources only on the necessary incremental data rather than re-transferring the entire VM image, thereby reducing replication time while maintaining complete data integrity.
3Reliability
If the VM is powered off before snapshot capture, then data loss is eliminated, but downtime increases
Solution Approach 1:
The system captures the initial snapshot while the virtual machine is still running, before initiating the power-off sequence. This preliminary action preserves the complete VM state including any in-memory data, ensuring no data loss occurs. The snapshot is then transferred to the target environment, and only after the source VM is powered off is the second incremental snapshot captured and applied, minimizing the time the application is unavailable.
Data Source
AI summary
Some users of a data management system (DMS) may use multiple computing environments to replicate and store virtual machines (VM)s, such as for backup and recovery purposes. For example, different replication environments may include one or more private data centers, one or more cloud environments or any combination thereof. A user may schedule a failover procedure for an application. A DMS may perform a failover procedure that reduces downtime and eliminates data loss. The DMS may capture and replicate a snapshot of a VM running on a source computing environment to a target computing environment, power down the VM on the source computing environment, capture and replicate a second snapshot of the VM to the target computing environment, and power on the VM at the target computing environment. As the additional snapshot includes a relatively small amount of data replication at the target computing environment may proceed quickly, reducing downtime.