ABSTRACT
The wide adoption of key-value (KV) data structures (such as in Apache Cassandra[1], Redis[2], RocksDB[3]) along with the ongoing efforts to support KV commands in the NVMe standard, motivates the development of KV at the drive level. The abstract discusses a development of a KV FTL (Flash Translation Layer) over Open Channel SSD that includes smart techniques for managing huge number of variable size objects with relatively low DDR footprint. The implementation is based on hardware and shows much higher throughput, improved latency, significant reduction in CPU load and high scalability compared to RocksDB, which is one of the most popular KV stores.
- 2019. cassandra. cassandra.apache.org/.Google Scholar
- 2019. redis. https://redis.io/.Google Scholar
- 2019. RocksDB. http://rocksdb.org/.Google Scholar
Recommendations
An efficient design and implementation of LSM-tree based key-value store on open-channel SSD
EuroSys '14: Proceedings of the Ninth European Conference on Computer SystemsVarious key-value (KV) stores are widely employed for data management to support Internet services as they offer higher efficiency, scalability, and availability than relational database systems. The log-structured merge tree (LSM-tree) based KV stores ...
A hybrid SSD with PRAM and NAND Flash memory
The speed of computing processor has been improved dramatically with multi-core architecture. However, the overall computer system performance shows slow improvement because of the sluggish speed of storage system. Several researches have been done to ...
An Embedded FTL for SSD RAID
DSD '15: Proceedings of the 2015 Euromicro Conference on Digital System DesignSolid State Disk (SSD) storage systems are the storage medium of choice in modern embedded devices, and so the performance, lifespan, and reliability of these devices is an increasing issue in many application domains. Previous work has proposed ...
Comments