Pseudo-Random Number Selection for Storage Device Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional testing devices struggle to mimic real-world scenarios effectively, particularly in generating workloads for storage devices that are random and unpredictable, while existing random number generation algorithms are cumbersome to modify to meet specific testing requirements.
Innovation Solution
A method and system for selecting numbers from multiple ranges using a non-transitory computer readable medium, which iteratively selects numbers to ensure uniqueness and randomness, utilizing a bit mapping mechanism to generate pseudo-random sequences for controlling aspects of network or storage device testing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional predefined workloads are used for testing, then testing can be performed with simple predefined scenarios, but the workloads cannot cover all real-world scenarios that customers may want to test
Solution Approach 1:
The patent segments the workload generation into multiple independent range definitions, each representing different workload characteristics. Instead of using a single complex predefined workload, the system divides the testing space into multiple ranges with different parameters (min/max values, weights, distributions), allowing comprehensive coverage through组合 of simpler segmented components.
Solution Approach 2:
The patent implements dynamic workload generation by allowing the testing system to adaptively select and combine different range definitions based on testing requirements. The workload characteristics can dynamically change between different range segments, enabling the system to cover diverse real-world scenarios rather than being stuck with static predefined workloads.
2Reliability
If random number generation algorithms are used to create unpredictable workloads, then random and unpredictable testing scenarios can be generated, but the algorithms cannot meet other requirements such as unique values and excluding or including certain values
Solution Approach 1:
The patent applies local quality by assigning different characteristics to different ranges. Each range can have its own minimum and maximum values, weightings, and distribution patterns. This allows the system to generate random values with specific local properties (e.g., certain ranges weighted higher, specific value exclusions) while maintaining overall randomness and unpredictability across the full testing space.
Solution Approach 2:
The patent changes parameters of the random number generation by allowing dynamic adjustment of range definitions, including min/max values, weights, and distributions. Instead of using a fixed random algorithm, the system modifies parameters to meet specific testing requirements while preserving randomness, such as adjusting range boundaries to exclude certain values or changing weightings to include specific patterns.
3Manufacturing precision
If existing random number generation algorithms are modified to meet specific testing requirements, then the algorithms can generate workloads with unique values and specific ranges, but the modification process is cumbersome and time intensive
Solution Approach 1:
The patent performs preliminary action by pre-defining multiple range configurations with different characteristics (min/max values, weights, distributions) before actual testing begins. These range definitions are prepared in advance and can be selectively combined during testing, eliminating the need for time-consuming modifications of random number generation algorithms during the testing setup phase.
Solution Approach 2:
The patent creates a universal range selection framework that can handle multiple testing requirements through a single unified mechanism. The same range definition structure supports various functionalities including unique value generation, value exclusions, different distributions, and weighted selections, replacing the need for multiple specialized algorithm modifications.
Data Source
AI summary
Methods, systems, and computer readable media for selecting numbers from multiple ranges are disclosed. One method includes receiving, information associated with a plurality of ranges, selecting, by a module implemented using a non-transitory computer readable medium, iteratively selecting numbers from within the ranges such that, during a selection iteration, a given number within one of the ranges is not selected more than once and such that a sequence of numbers selected during the selection iteration appears to be random, and utilizing the numbers selected during the selection iteration to control at least one aspect of testing a network or storage device.