Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for grouping LSM tree indexes based on GPU

A technology of LSM tree and index, applied in the field of GPU database, can solve the problem of slow query speed of LSM tree, and achieve the effect of improving query efficiency

Active Publication Date: 2020-11-27
NORTHEASTERN UNIV
View PDF3 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For the slow query speed of the LSM tree, the present invention eliminates a large number of unnecessary queries through the Bloom filter, thereby improving the query efficiency

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for grouping LSM tree indexes based on GPU
  • Method for grouping LSM tree indexes based on GPU
  • Method for grouping LSM tree indexes based on GPU

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] The specific embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings.

[0034] The present invention is divided into two parts: the insertion, update and deletion of the GPU grouped LSM tree, and the query on the GPU grouped LSM tree,

[0035] For insert update or delete operations, firstly, the data needs to be separated from the Key and Value in the memory, the Value is stored in the memory, and the address of the Value is recorded in the Hash table in the memory at the same time; then the data is copied from the memory To the buffer of the GPU global memory, the Key and the corresponding Value address will be copied here. Then the GPU will sort the data in the buffer once. After sorting, it will be copied to the first layer of the GPU grouped LSM tree. If the first layer cannot fit, a merge operation will be triggered and the data will be brushed to the lower layer. .

[0036] For the query operation, in ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a grouping LSM tree indexing method based on a GPU, and relates to the technical field of GPU databases. According to the method, firstly, data is preprocessed, when value is lengthened, cache cannot be well utilized when query is carried out on a GPU, and the data transmission cost can be increased. Aiming at the conditions, the Key and the Value in the data are separated,only the address of the Value is stored in the GPU, and the real Value is stored in the memory. Aiming at the problem of low LSM insertion speed, each layer of an original LSM tree is divided into a plurality of groups, each group is an ordered array, and parallel merging is performed through a large number of threads on a GPU when the groups are merged to the next layer. Because the LSM trees aregrouped, the query needs to spend a higher cost. In order to improve the query speed, a Bloom filter suitable for a GPU structure is designed on the GPU, and a large amount of unnecessary query expenditure is reduced through the Bloom filter.

Description

technical field [0001] The invention relates to the technical field of GPU databases, in particular to a method for grouping LSM tree indexes based on GPUs. Background technique [0002] With the advent of the era of big data, the total amount of data and the amount of data access are increasing explosively. Traditional relational databases are no longer able to meet such high-concurrency access scenarios. The Nosql database does not depend on the traditional structure of the relational database and is more flexible and convenient, so it is very suitable for cloud storage, e-commerce and web access, etc. Key-Value (KV) is the basic type of Nosql database, and a large amount of unstructured data can be read and written through simple interfaces such as GET, PUT, and DELETE. [0003] The index structure of the current mainstream Nosql database is a log-structured merge tree (log-sturctre mergetree, LSM Tree). LSM tree is one of the important data structures. It is widely u...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/901G06F16/903G06F16/245
CPCG06F16/24569G06F16/9027G06F16/90335
Inventor 谷峪李万李传文李芳芳于戈
Owner NORTHEASTERN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com