Cascade Ordering for Storage Backup Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing multiple target flashcopy implementations do not effectively minimize read access to the production disk during background copy processes, especially in cascaded configurations, leading to increased time and inefficiency with the number of copies required.

Innovation Solution

A method where backup processes are initiated from a single source storage volume to multiple target storage volumes, with each target added to the cascade in inverse proportion to its copy rate, and started in the reverse order of addition, allowing each map to have its own background copy rate without waiting for others to complete.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multiple backup processes are initiated simultaneously to multiple target storage volumes, then the backup operation can be performed in parallel improving productivity, but the read access to the production disk increases leading to increased time and inefficiency

Engineering Contradiction:
Improvebackup operation speedVSAvoidread access time to production disk
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent inverts the traditional backup initiation approach by starting backup processes in reverse order of their copy rates. Instead of starting all backups simultaneously or in arbitrary order, the system starts the slowest backup process first, then the next slowest, and so on. This inversion ensures that the production disk is accessed for reading in an optimized sequence that minimizes total read time while still enabling parallel backup operations to multiple targets.

Inventive Principle:
Principle #13The other way round (Inversion)

2Reliability

If cascaded flashcopy configurations are used to create multiple copies, then the system can maintain data redundancy across multiple targets, but the complexity of managing multiple cascaded maps increases

Engineering Contradiction:
Improvedata redundancyVSAvoidcascade management complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the parameter of backup process initiation timing based on copy rate characteristics. By dynamically adjusting when each backup process starts according to its specific copy rate, the system simplifies the management of cascaded configurations. This parameter-based approach allows the system to handle multiple cascaded maps more efficiently by automatically optimizing the startup sequence rather than requiring complex manual management of cascade relationships.

Inventive Principle:
Principle #35Parameter changes

3Ease of manufacture

If each target storage volume is added to the cascade in arbitrary order, then the system implementation is simpler, but the number of read operations on the production disk increases

Engineering Contradiction:
Improvesystem implementation easeVSAvoidbackup efficiency
Core Design Contradiction:
Ease of manufactureVSProductivity

Solution Approach 1:

The patent optimizes the cascade construction by changing the order parameter in which target storage volumes are added to the cascade. Specifically, targets are added in order of their copy rates (slowest first, fastest last), which minimizes the number of read operations required on the production disk. This parameter optimization maintains reasonable implementation complexity while significantly improving backup efficiency by reducing redundant reads from the production disk.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS9063894B2Cascade ordering
Publication Date: 2015.06.23 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US9063894B2 patent drawing
  • US9063894B2 patent drawing
  • US9063894B2 patent drawing

AI summary

A method comprises receiving one or more instructions initiating a plurality of backup processes from a single source storage volume to a plurality of target storage volumes, adding each target storage volume to a cascade of target storage volumes from the source storage volume, the target storage volumes added to the cascade in an order inversely proportional to the copy rate of the respective backup process, and starting each backup process in turn, the backup processes started in an order from the most recent target storage volume added to cascade to the first target storage volume added to cascade.