Clustered Database Instant Restore via Flashcopy and Preflight Checks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional copy data management techniques are inadequate for instant restore of Very Large Databases (VLDBs) and provisioning of clustered databases with near zero downtime, especially in scenarios involving Out-of-Band protection and clustered database environments.
Innovation Solution
A computerized method and system for restoring a clustered database with near zero downtime by performing a preflight check on nodes, creating a flashcopy of the backup image, mapping it to each node, and switching the database to run from the flashcopy, thereby ensuring minimal downtime during the restoration process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional filesystem-based backup mounting is used, then instant data availability is achieved, but the method is inadequate for VLDB instant restore and clustered database provisioning
Solution Approach 1:
The patent uses flash copy technology to create a point-in-time copy of the backup image that can be instantly mounted and switched to without requiring data migration or processing. This copy operation enables instant restore capability for VLDBs and clustered databases while maintaining the reliability of traditional backup methods.
Solution Approach 2:
The system performs preflight checks before the restore operation to validate user permissions, check software locations, verify running processes, and ensure database status. These preliminary actions ensure that the restore operation can proceed instantly without encountering blocking issues, thereby achieving reliable instant restore for complex database environments.
2Reliability
If preflight checks are performed to validate user permissions and software status, then restore reliability is improved, but restore time is increased
Solution Approach 1:
The preflight checks are performed as preliminary actions before the actual restore operation begins. By validating user permissions, checking software locations, verifying running processes, and confirming database status in advance, the system ensures that the subsequent restore operation can proceed instantly without time-consuming validation steps during the restore itself.
Solution Approach 2:
The preflight checks act as a cushioning mechanism that identifies and resolves potential issues before they can block the restore operation. This beforehand validation ensures that even though checks are performed, they do not result in restore failures or delays during the actual restore process.
3Loss of time
If flashcopy is created and mapped to multiple nodes, then clustered database restoration with near zero downtime is achieved, but system complexity increases
Solution Approach 1:
The flash copy creates a point-in-time copy of the backup image that can be simultaneously mapped to multiple cluster nodes. This copying mechanism enables near zero downtime restoration by allowing all nodes to access the restored data instantly without requiring sequential restoration or data migration, thereby achieving minimal downtime while managing complexity through automated operations.
Solution Approach 2:
The system merges the flash copy creation and multi-node mapping operations into a single coordinated process. By combining these operations, the system achieves near zero downtime restoration for clustered databases while reducing the overall complexity compared to performing each operation separately.
Data Source
AI summary
Restoring a clustered database having a plurality of nodes each having database from a failed storage device by receiving a request to restore a backup image of a failed shared storage device associated with the clustered database to a time; performing a preflight check including at least one checklist process; terminating the restore when any checklist process fails; when each checklist process succeeds completing the restore by creating at least one flashcopy associated with the backup image, mapping to each of the plurality of nodes an associated portion of the at least one flashcopy, mounting the at least one flashcopy to the node as a diskgroup, and switching the clustered database to run from the diskgroup.