ASM Diskgroup Backup for Instant VLDB Restore
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional copy data management techniques are inadequate for instant restore of Very Large Databases (VLDBs) when production databases go offline and for provisioning clustered databases in test and development environments, particularly for Oracle Real Application Clusters (RAC), as they often result in significant downtime and increased operational complexity.
Innovation Solution
The implementation of Automatic Storage Management (ASM) diskgroup based backup systems that enable incremental forever backups, providing resource-efficient and high-throughput backup solutions for clustered databases, allowing for instant restore and near-zero downtime data migration, as well as near-constant time provisioning of test and development environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional filesystem-based backup methods are used for VLDB, then backup can be performed, but restore time is excessively long and downtime is significant
Solution Approach 1:
The patent creates a copy of the production database backup to a test node, allowing instant restoration without lengthy traditional restore processes. The backup copy is mounted as a disk group on the test node, enabling immediate data access and database operations.
Solution Approach 2:
The patent performs preliminary actions by pre-configuring the backup environment with disk groups and mounting structures before restoration is needed. The backup database is prepared in advance on the test node, so when restoration is required, the database can be activated immediately without setup delays.
2Productivity
If traditional backup methods are used for clustered databases, then backup can be performed, but provisioning time for test and development environments is excessive
Solution Approach 1:
The patent copies the production clustered database backup to a test node and mounts it as a disk group, enabling instant provisioning of test and development environments. This eliminates the need to recreate database structures and load data sequentially.
Solution Approach 2:
The patent segments the database provisioning process by separating the backup creation (on production node) from the database activation (on test node). This allows parallel processing where the backup is prepared independently and then instantly activated on the target node.
3Reliability
If out-of-band backup is performed for clustered databases, then backup coverage is improved, but backup complexity and operational overhead increase
Solution Approach 1:
The patent introduces an intermediary backup node that receives backups from the production node and serves restoration requests to test nodes. This intermediary structure simplifies the out-of-band backup process by centralizing backup management and eliminating direct complex interactions between production and test environments.
4Ease of operation
If instant restore of VLDB is required, then data availability is improved, but traditional methods cause significant downtime
Solution Approach 1:
The patent performs preliminary restoration actions on a test node before they are needed. The backup is mounted and the database structure is prepared in advance, so when instant restore is required, only a simple activation step is needed, achieving near-zero downtime.
Solution Approach 2:
The patent creates a ready-to-use copy of the database on a test node with identical structure and data. When instant restore is needed, the system switches to the pre-prepared copy, eliminating the lengthy traditional restore process and achieving immediate data availability.
Data Source
AI summary
Systems and methods for backing up clustered databases such that production backups associated with the clustered databases are restorable with near zero downtime. First data indicative of a backup request of a clustered database is transmitted by a first computing device to a second computing device. The first computing device maps the backup staging disks to each of a plurality of nodes. The second computing device adds the mapped staging disks to a backup staging diskgroup on a primary node. The second computing device mounts the backup staging diskgroup to each of the plurality of nodes. The second computing device executes a full backup from at least one of the plurality of nodes to the backup staging diskgroup. The second computing device transmits a catalog of backup information to a database associated with the first computing device. The first computing device creates a snapshot of the first backup staging disk.