A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf
.
Filters
Efficient Compactions Between Storage Tiers with PrismDB
[article]
2022
arXiv
pre-print
Our key contribution is how to efficiently migrate and compact data between two different storage tiers. ...
with the cost of compaction I/O in slow storage. ...
A model and algorithm for efficient compactions for multi-tiered storage that balances data placement with compaction I/O in the slow tier. (3) Popularity scoring. ...
arXiv:2008.02352v6
fatcat:47hxkljotjbabklbgff4ww7efy
HotRAP: Hot Record Retention and Promotion for LSM-trees with tiered storage
[article]
2024
arXiv
pre-print
Our experiments show that HotRAP outperforms state-of-the-art LSM-trees on tiered storage by up to 3.3× compared to the second best for read-only and read-write-balanced workloads with common access skew ...
However, frequently accessed records may have been compacted and reside in slow storage, and existing algorithms are inefficient in promoting these "hot" records to fast storage, leading to compromised ...
Compared to the state-of-the-art LSM-trees on tiered storage (e.g., RocksDB with secondary cache [26] , Mutant [36] , and PrismDB [29] ), HotRAP achieves 3.3× speedup over the second best for read-only ...
arXiv:2402.02070v1
fatcat:aez2dytcerh6javaxllgatwh24
Efficient LSM-Tree Key-Value Data Management on Hybrid SSD/HDD Zoned Storage
[article]
2022
arXiv
pre-print
storage. ...
HHZS leverages hints issued by the flushing, compaction, and caching operations of the LSM-tree KV store to manage KV objects in placement, migration, and caching in hybrid ZNS SSD and HM-SMR HDD zoned ...
Hybrid zoned storage incurs additional data movement between heterogeneous zoned storage devices (i.e., ZNS SSDs and HM-SMR HDDs). ...
arXiv:2205.11753v1
fatcat:fefuu3zh3jgt3d46feo4vqffda
Building A Fast and Efficient LSM-tree Store by Integrating Local Storage with Cloud Storage
2022
ACM Transactions on Architecture and Code Optimization (TACO)
Designing an LSM store by integrating local storage with cloud storage services is a promising way to balance the cost and performance. ...
To improve metadata space-efficiency and read performance, RocksMash uses an LSM-aware persistent cache that stores metadata in a space-efficient way and stores popular data blocks by using compaction-aware ...
Taking these findings into consideration, we have presented RocksMash, a fast and efficient LSM store that efficiently splits LSM-tree between local storage and cloud storage to achieve cost-effectiveness ...
doi:10.1145/3527452
fatcat:hupcq4bgv5agvmhtti22u2i6tm
Key-Value Stores on Flash Storage Devices: A Survey
[article]
2022
arXiv
pre-print
Key-value stores (KV) have become one of the main components of the modern storage and data processing system stack. ...
With the increasing need for timely data analysis, performance becomes more and more critical. In the past, these stores were frequently optimised to run on HDD and DRAM devices. ...
At the same time, it comes with its own set of challenges. Such as efficient translations between data structures. ...
arXiv:2205.07975v1
fatcat:5sv2j2m3l5hrlnpjohlcxsh574
SpanDB: A Fast, Cost-Effective LSM-tree Based KV Store on Hybrid Storage
2021
USENIX Conference on File and Storage Technologies
Compared with KVell, a system designed for high-end SSDs, SpanDB achieves 96-140% of its throughput, with a 2.3-21.6⇥ lower latency, at a cheaper storage configuration. ...
with polling-based I/O. ...
Related Work Tiered storage. ...
dblp:conf/fast/ChenRLMX21
fatcat:qu2lmqp56fcgfliutfayre3oia