Persistent Volume Mapping for Container Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In container orchestration environments, the mapping information between dynamically created volumes and declaratively described volume claims is lost during restoration, leading to the creation of new volumes instead of reusing existing ones after a planned outage or failure, which prevents exact recreation of the previous environment state.
Innovation Solution
A method and system for managing volume management information that maps volume claim information with a logical volume identifier associated with persistent volume information, allowing for the retrieval and provision of this information to a volume provisioning function to recreate existing volumes during recovery, ensuring that the same environment can be restored as before the outage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dynamic provisioning is performed during initial setup, then storage volumes are automatically created and mapped to volume claims, but after restoration from outage the mapping information is lost and new volumes are created instead of reusing existing ones
Solution Approach 1:
The system performs preliminary actions by storing volume management information (mapping between volume claims and persistent volumes) in a persistent storage location before the outage occurs. During recovery, this pre-stored information is retrieved and used to restore the original volume mappings, preventing the loss of mapping information that would otherwise occur during restoration.
Solution Approach 2:
The system creates a copy of the volume management information and stores it separately from the container orchestration environment metadata. This copy persists through outages and restorations, allowing the original volume mappings to be reconstructed after recovery without relying on the transient dynamic provisioning process.
2Ease of repair
If backup processes restore metadata after shutdown, then the environment can be recovered, but the volume mapping information is reverted to backup frequency intervals and cannot reproduce the exact pre-shutdown state
Solution Approach 1:
The system extracts the volume management information from the container orchestration environment metadata and stores it separately in a persistent storage location. This separation allows the volume mapping information to be preserved independently of the backup frequency of the main environment metadata, enabling exact pre-shutdown state reproduction without being constrained by backup intervals.
3Productivity
If new volumes are created during recovery, then the environment can be restored, but the existing volume mappings are lost and new volumes consume additional storage resources
Solution Approach 1:
The system implements a feedback mechanism where volume provisioning requests are checked against the stored volume management information before creating new volumes. If existing volume mappings are found in the stored information, the system retrieves and reuses those existing volumes instead of creating new ones, thereby preventing storage resource waste while maintaining fast recovery through automated decision-making.
Data Source
AI summary
Disclosed herein is a technique to automatically re-bind applications and storage volumes during recovery from planned outage or platform failure and disaster recovery. Such implementations can involve managing volume management information that maps volume claim information with a logical volume identifier associated with persistent volume information; and for receipt of a request for a new volume, referencing the volume management information to retrieve an associated logical volume identifier and the persistent volume information based on the volume claim information associated with the request; and providing the retrieved persistent volume information to a volume provisioning function.