HBase secondary full-text indexing method and system based on phoenix

Abstract

The invention proposes a phoenix-based HBase secondary full-text indexing method and system, comprising the steps of: establishing a secondary full-text index table for the HBase original data table, storing it in the Elasticsearch storage engine, and establishing a relationship between the secondary full-text index table and the HBase original data table. Associate foreign keys, so that after the full-text retrieval conditions are pushed down to the Elasticsearch storage engine, the query result of the full-text index table of the Elasticsearch storage engine is the rowkey of the original data table of HBase; when performing full-text indexing, the phoenix query server converts the client query statement to Elasticsearch storage The engine can execute the query expression, and push down the Elasticsearch storage engine query; the phoenix query server uses the query result returned by the Elasticsearch storage engine as the RowKey collection of the Hbase original data table to query in the Hbase storage engine, and returns the final result of the query to the client. The invention converts the secondary index function into a full-text index based on ElasticSearch and a RowKey query based on HBase, expands the original Phoenix function, and meets the needs of users for full-text search.

Description

technical field [0001] The invention relates to the technical field of database indexing, in particular to a phoenix-based Hbase secondary full-text indexing method and system. Background technique [0002] The advantage of HBase is RowKey query and range query, while non-RowKey query will become extremely slow. For this feature, phoenix provides a secondary index function based on HBase. Phoenix implements the secondary index function by creating two HBase tables, that is, an index table index table1 is created when the index is built, and the RowKey column of the index table is the non-RowKey column of the original table. However, this secondary indexing technique has the following drawbacks: [0003] Since both the original data table and the index table are stored in HBase, the query of the index table is also required to be a RowKey query, and the query of the full-text index of the non-RowKey column of the original table cannot be satisfied. Contents of the inventio...

AI Technical Summary

Problems solved by technology

[0003] Since both the original data table and the index table are stored in HBase, the query of the index table is also required to be a RowKey query, which cannot be satisfied for the query of the full-text index of the non-RowKey column of the original table.

Images