Tag-Based Resource Mapping for Virtual Machine Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In disaster recovery management for virtual machines, existing methods face challenges in maintaining consistent host-storage unit topology during failover due to practical limits on the number of mounted storage units and differences in capacity between production and recovery sites, leading to inefficient processes and potential virtual machine degradation or failure.
Innovation Solution
The solution involves using a resource-consumer mapping process that assigns tags to hosts and datastores, allowing for a flexible and efficient mapping of datastores to hosts at the recovery site, maintaining logical divisions and avoiding the need for mounting every storage unit on every host, thereby enabling failover without exceeding storage unit limits and preserving initial site configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If every recovered storage unit is mounted on every host at the recovery site, then complete storage availability is achieved, but the number of mounted storage units exceeds practical limits on hosts
Solution Approach 1:
The patent segments the host-storage unit mapping by creating separate mapping relationships for production site and recovery site. Instead of mounting every storage unit on every host, it divides the mapping into: (1) production site topology mapping stored in a database, and (2) recovery site topology mapping created during failover. This segmentation allows selective mounting of only necessary storage units on recovery hosts, avoiding the exponential growth of mounted storage units.
Solution Approach 2:
The patent performs preliminary action by capturing and storing the production site host-storage unit topology mapping in a database before failover occurs. This pre-captured topology information is then used during failover to automatically recreate the appropriate mapping at the recovery site, eliminating the need for manual configuration and avoiding the need to mount all storage units on all hosts.
2Stability of the object's composition
If storage units are mapped identically at the recovery site as at the protected site, then topology consistency is maintained, but the different capacities between sites make one-to-one matching impractical
Solution Approach 1:
The patent applies parameter changes by allowing the recovery site topology mapping to differ from the production site mapping in terms of specific host-storage unit assignments, while maintaining the logical topology relationships. The system changes the physical parameters (which specific hosts mount which storage units) to accommodate different site capacities, while preserving the logical topology parameters (which virtual machines access which storage units). This enables adaptation to different site capacities while maintaining topology consistency at the logical level.
3Ease of operation
If manual mapping of storage units to hosts is performed at the recovery site, then precise control over mapping is achieved, but the process becomes inefficient and time-consuming
Solution Approach 1:
The patent implements self-service by enabling automatic recreation of host-storage unit topology mapping at the recovery site. The system automatically retrieves the production site topology from the database, translates it to the recovery site context, and configures the mapping without human intervention. This self-service automation maintains precise control over the mapping (ensuring correctness) while dramatically improving failover speed and efficiency by eliminating manual configuration steps.
Data Source
AI summary
Mapping computer resources to consumers in a computer system is described. In an example, a method of mapping computer resources to consumers in a computer system includes: receiving tags assigned to the computer resources at a resource manager executing in the computer system, where the resource manager: identifies a first tag assigned to a first computer resource; determines whether a first consumer is associated with the first tag; enables the first consumer to access the first computer resource if the first consumer is associated with the first tag; and prevents the first consumer from accessing the first computer resource if the first consumer is not associated with the first tag.