Dynamic Journal Disk Assignment in Storage Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing content address storage system experiences decreased writing performance and reliability due to contention between fragment data and journal storage on disks with OS regions, leading to delayed data block completion and potential data loss in case of disk failure.
Innovation Solution
A storage system is configured to separately store journals and fragment data on different disks, with a dynamic setting mechanism that adjusts the disk roles based on available capacity, ensuring that journals are not fixedly stored on a single disk and are replicated across multiple disks for redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If journals are fixedly stored on disks with OS regions, then redundancy is maintained, but writing performance decreases due to contention between fragment data and journal storage
Solution Approach 1:
The patent applies dynamics by making the journal storage location flexible rather than fixed. The system dynamically selects which disk will store the journal based on real-time conditions such as disk capacity and current journal location, allowing the journal storage assignment to adapt and change over time rather than remaining static on specific disks with OS regions.
Solution Approach 2:
The patent segments the storage system by separating journal storage from fragment data storage. By designating specific disks as journal storage disks and others as fragment storage disks, the system divides the storage functionality into distinct segments, eliminating the contention that occurs when both journal and fragment data are stored on the same disk.
2Quantity of substance
If journals are stored on the same disks as fragment data, then storage capacity is maximized, but reliability decreases due to potential simultaneous loss of journal and fragment data
Solution Approach 1:
The patent segments storage disks into two functional groups: journal storage disks and fragment storage disks. This segmentation ensures that journals and fragment data are stored on physically separate disks, so that if one disk fails, the other remains intact and functional, preventing simultaneous loss of both journal and fragment data.
Solution Approach 2:
The system introduces an intermediary mechanism - the journal disk assignment logic - that acts as a mediator between the need for storage capacity and the need for data safety. This intermediary ensures that journal storage is always separated from fragment data storage, maintaining reliability while still utilizing available storage capacity across multiple disks.
3Quantity of substance
If all disks are used for data storage including those with OS regions, then storage efficiency increases, but writing performance decreases due to process contention
Solution Approach 1:
The patent segments the disk pool by designating specific disks as journal storage disks and others as fragment storage disks. This segmentation allows all disks to be utilized for storage purposes (maintaining storage efficiency) while preventing process contention by ensuring that journal writing operations and fragment data writing operations occur on separate disks.
Solution Approach 2:
The patent extracts the journal storage function from the general data storage function. By taking out the journal storage requirement and assigning it to specific designated disks separate from fragment data storage, the system eliminates the harmful interaction and process contention that would occur if both functions shared the same disk resources.
Data Source
AI summary
A storage system includes a plurality of storing unit and a data processing unit configured to store data into the plurality of storing unit. The data processing unit includes: a storage destination setting unit configured to set a journal storing unit configured to store a journal showing a data processing status of the storage system from among the plurality of storing unit, and set the plurality of storing unit other than the set journal storing unit as fragment storing unit configured to distributedly store a plurality of fragment data forming storage target data, respectively; and a distribution storage controlling unit configured to store the journal into the storing unit set as the journal storing unit by the storage destination setting unit, and distribute and store the plurality of fragment data into the plurality of storing unit set as the fragment storing unit, respectively.