Cluster Live Mounts for Instant Database Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database recovery methods are slow and inefficient, requiring the migration of large files between a downed production host and a backup device, leading to lengthy recovery times and significant network overhead, which is unacceptable for critical databases that cannot afford prolonged downtime.
Innovation Solution
A cluster-based database system implements instant recovery by live mounting directories to multiple nodes in a distributed database cluster, allowing each node to manage directories and files, eliminating the need for large file migration by using Network File System (NFS) mounts, and utilizing a remote connector agent to execute pre-configured scripts for rapid recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional database recovery methods are used to migrate large files between downed production host and backup device, then database recovery can be performed, but recovery time becomes lengthy and network overhead increases significantly
Solution Approach 1:
The patent segments the database files into smaller chunks and distributes them across multiple nodes in a cluster. Instead of migrating entire large files sequentially, the system divides the recovery process into parallel operations where multiple file segments can be transferred and restored simultaneously to different nodes, significantly reducing total recovery time.
Solution Approach 2:
The patent transitions from a single-node recovery model to a multi-node distributed cluster model. By adding the dimension of parallel processing across multiple nodes, the system can perform recovery operations concurrently rather than sequentially, transforming a time-consuming sequential file migration into a parallel restoration process that dramatically reduces recovery time.
2Reliability
If conventional database recovery methods are used to migrate large files, then database recovery can be performed, but network overhead becomes large
Solution Approach 1:
The patent segments large database files into smaller manageable chunks that can be distributed across multiple nodes. This segmentation allows the network to transfer data in smaller, more efficient packets rather than moving entire large files back and forth, reducing total network traffic and overhead during the recovery process.
Solution Approach 2:
The patent merges multiple recovery operations into a single coordinated cluster process. Instead of each node performing separate file migration operations that generate cumulative network overhead, the system combines these operations into a unified recovery mechanism where nodes work in parallel on different segments, reducing redundant network communication and overall overhead.
3Reliability
If conventional database recovery methods are used, then database files can be restored, but recovery process is slow and creates lengthy downtime
Solution Approach 1:
The patent divides the database restoration process into parallel segments that can be executed simultaneously across multiple cluster nodes. Instead of restoring files sequentially one after another, the system segments the restoration task so that multiple nodes can work on different file segments concurrently, dramatically increasing recovery speed and reducing downtime.
Solution Approach 2:
The patent enables continuous parallel restoration operations across multiple nodes simultaneously. Rather than waiting for one node to complete its file restoration before starting the next node, the system maintains continuous useful action by having multiple nodes restore different file segments concurrently, maximizing recovery throughput and minimizing total downtime.
Data Source
AI summary
A database can be instantly recovered by a cluster mapped to the database. Nodes of the cluster are mapped over channels to directories of the database. Scripts are generated from one or more templates that specify the order and values to be executed to perform a database job, such as database recovery. To recover the database, a template is executed that generates and populates scripts, which are processed on the host of the database to recover the database in a nearly instant manner without transferring data files.