Storage Subsystem Journal Archiving to Tape
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional disk-based Continuous Data Protection (CDP) methods face challenges with excessive capacity and cost due to continuous journal data acquisition, leading to limited retention of long-term data, and tape-based methods burden the host with high load and inefficient restore processes.
Innovation Solution
A storage system utilizing a combination of random and sequential storage devices, with a backup unit and restore unit, where journal data is written to a disk and then copied to tape, allowing for low-cost journal retention and efficient restore without host load, using a controller program to manage journal pools and tape groups for efficient data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If journal data is continuously acquired and stored on disks for CDP, then data protection capability is improved, but storage capacity cost increases excessively
Solution Approach 1:
The storage system is segmented into two distinct parts: disk storage for high-speed random access journal data and tape storage for low-cost sequential archive data. This segmentation allows the system to maintain data protection capabilities while reducing overall storage costs by placing different types of data on different media optimized for their respective access patterns.
Solution Approach 2:
The storage subsystem acts as an intermediary between the host and the tape archive. The subsystem's controller manages journal acquisition from the host, stores it temporarily on disks, and handles the archiving to tape, thereby protecting the host from direct tape I/O operations and enabling efficient data management without host involvement.
2Quantity of substance
If journal data is stored on tapes by host, then storage capacity is reduced, but host performance deteriorates due to high load
Solution Approach 1:
The storage subsystem serves as an intermediary that assumes the burden of tape I/O operations. The controller directly manages journal acquisition, formatting, and archiving to tape, completely isolating the host from the high load of sequential tape operations while maintaining efficient data protection capabilities.
Solution Approach 2:
The storage subsystem is designed to be self-sufficient in managing journal data archiving. The controller autonomously handles all aspects of journal acquisition, formatting with sequence information, and tape archiving without requiring host CPU or I/O resources, thereby preserving host performance while achieving cost-effective storage.
3Quantity of substance
If tape-based journal storage is used, then storage cost is reduced, but restore efficiency deteriorates
Solution Approach 1:
The system segments restore operations into two paths: fast restores from disk for recent data and slower restores from tape for historical data. The controller intelligently determines the source based on the required restore point, optimizing restore efficiency by utilizing the appropriate storage medium for each scenario.
Solution Approach 2:
Journal data is pre-formatted with sequence information and metadata during the archiving process, enabling efficient random access and selective restore operations. This preliminary organization of data allows the restore unit to quickly locate and retrieve specific journal entries without scanning entire tape sequences, significantly improving restore efficiency.
Data Source
AI summary
One or more random access-type storage devices comprise a volume group and a journal volume. The volume group contains a volume pair configured from a primary volume and a secondary volume. Whenever data is received from a host device, this received data is written in the primary volume, a journal containing the write sequence of the data and the data is produced, and the journal is written in a journal volume. A plurality of journals accumulated in a journal volume are written in a sequential access-type storage device. Based on the write sequence recorded in each of the plurality of journals recorded in the sequential access-type storage device, the data of the journals is restored to the primary volume for storing the data and to a secondary volume from which the volume pair is configured.