Bloom filter-based data query system

Abstract

The invention relates to a Bloom filter-based data query system, wherein the system comprises an agent layer, a first query channel and a second query channel; the agent layer is connected with the first query channel and the second query channel, is used for receiving a data query request sent by a client, distributing the data query request according to a preset query strategy, and receiving a first query result returned by the first query channel and / or a second query result returned by the second query channel, feeding back a query result to the client based on the query strategy, the first query result and / or the second query result; the first query channel comprises a cache layer and a persistence layer; the second query channel comprises a Bloom filter layer. According to the system, the Bloom filter layer can be additionally arranged on the premise of not modifying an original database architecture code, and the data query efficiency can be improved.

Description

technical field [0001] The invention relates to the technical field of databases, in particular to a data query system based on a Bloom filter. Background technique [0002] In the existing database architecture, a multi-layer caching mechanism is usually adopted, and the database with high reliability, slow speed and large capacity is used as the data persistence layer; the database with low reliability, high speed and small capacity is used as the data cache layer. Only hot data is stored in the cache layer, and full data is stored in the persistence layer. When performing data query tasks, such as querying by keyword (Key), first query the cache layer with faster speed. If the corresponding value (Value) can be found, the query task is successful, otherwise, it needs to go to the persistence layer. Query again, and cache penetration occurs. When cache penetration occurs, the execution time of the query task will be longer, and the data query efficiency will be low. [...

AI Technical Summary

Problems solved by technology

However, for the already designed database architecture, it is necessary to modify the code to add the Bloom filter layer, and modifying the code will bring vulnerability risks

In addition, when the amount of data in the database is large, the corresponding Bloom filter layer will occupy a large amount of memory, resulting in greater pressure on the stand-alone server storing the filter, which will still affect the efficiency of data query

Images