An Effective Memory-Mapped Key-Value Store for SSD
ANIL AHIR
, JITENDRA MADAVI
parallelism, SSD, B-Trees, synchronous, asynchronous
Power consumption and space complexity which leads CPU overheads have to limit for efficient data processing work. Here, we recommend a key-value storage that aims servers with SSD storage, where CPU overhead and Input/output strengthening are more important blocks compared to Input/output arbitrariness. Here we seen 2 sources of overhead in key-value stores are: First- the usage of an Input/output cache memory to access devices, which suffers overhead even for data that reside in memory and Second- The use of compaction in Log Structure Merge-Trees that continually perform merging and sorting of huge data values. To escape this possibilities our data will move from one level to other by executing fractional instead of complete data restructuring via the usage of a full index per-level. In addition, we suggest the uses memory-mapped I/O via a custom kernel path with Copy-On-Write. Here we try to implement new methods to modify memory mapped Input/output in Linux system.
"An Effective Memory-Mapped Key-Value Store for SSD", IJSDR - International Journal of Scientific Development and Research (www.IJSDR.org), ISSN:2455-2631, Vol.7, Issue 9, page no.494 - 495, September-2022, Available :https://ijsdr.org/papers/IJSDR2209084.pdf
Volume 7
Issue 9,
September-2022
Pages : 494 - 495
Paper Reg. ID: IJSDR_201710
Published Paper Id: IJSDR2209084
Downloads: 000347294
Research Area: Computer Science & Technology
Country: Mumbai, MAHARASHTRA, India
ISSN: 2455-2631 | IMPACT FACTOR: 9.15 Calculated By Google Scholar | ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 9.15 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: IJSDR(IJ Publication) Janvi Wave
