ACM Transactions on Database Systems (TODS)
An efficient management scheme for large-scale flash-memory storage systems
Proceedings of the 2004 ACM symposium on Applied computing
An integrated approach to recovery and high availability in an updatable, distributed data warehouse
VLDB '06 Proceedings of the 32nd international conference on Very large data bases
A flash-memory based file system
TCON'95 Proceedings of the USENIX 1995 Technical Conference Proceedings
Incrementally parallelizing database transactions with thread-level speculation
ACM Transactions on Computer Systems (TOCS)
Transaction Management for Flash Media Databases in Portable Computing Environments
Journal of Intelligent Information Systems
Improving Flash Wear-Leveling by Proactively Moving Static Data
IEEE Transactions on Computers
Aether: a scalable approach to logging
Proceedings of the VLDB Endowment
Adaptive logging for mobile device
Proceedings of the VLDB Endowment
Sector log: fine-grained storage management for solid state drives
Proceedings of the 2011 ACM Symposium on Applied Computing
Transactional In-Page Logging for multiversion read consistency and recovery
ICDE '11 Proceedings of the 2011 IEEE 27th International Conference on Data Engineering
A survey of B-tree logging and recovery techniques
ACM Transactions on Database Systems (TODS)
Scalability of write-ahead logging on multicore and multisocket hardware
The VLDB Journal — The International Journal on Very Large Data Bases
LogBase: a scalable log-structured database system in the cloud
Proceedings of the VLDB Endowment
Flag Commit: Supporting Efficient Transaction Recovery in Flash-Based DBMSs
IEEE Transactions on Knowledge and Data Engineering
Hi-index | 0.00 |
This research is motivated by the needs for data manipulation over flash-memory storage systems, where flash memory has become a popular data storage medium for mobile devices. Different from the past work, we are interested in recovery designs with steal and no-force policies over ARIES-based recovery, that is widely adopted in many database systems. In this paper, we aim at the proposing of a design that improves not only the logging overheads but also the recovery performance for DBMSs over MLC (and SLC) flash memory. A logging mechanism with start, update and end log blocks are proposed to support efficient transaction recovery in joint considerations of both wear-leveling and garbage collection over flash memory. A recovery mechanism is proposed to efficiently find out the transactions that need to be redone or undone to address the steal and no-force polices. The capability of the proposed logging and recovery methodology was evaluated by extensive experiments, in which significant improvement on recovery performance was achieved, compared with the past work.