H base one page
0 likes289 views
HBase is a NoSQL database that stores data in labeled tables with rows containing a sortable key and an arbitrary number of columns. It provides scalability, automatic failover, distributed regions, integration with Hadoop/HDFS, and APIs for access. HBase is suitable for large datasets, variable schemas, and key-based access without complex queries.
H base one page
- 1. HBase is a NoSQL (Not Only SQL) database. HBase store data rows in labeled tables, each row have sortable key and an arbitrary number of columns. v HBase is linear and modular scaling v It have automatic region server failover v Tables are distributed on the cluster via regions which supports automatic sharding v Hadoop/HDFS integration v Supports MPP (massively parallelized processing) v Support Thrift and REST API v Supports Block Cache and Bloom Filters for high volume query optimization v Provides build-in web pages for operational insight v Supports strongly consistent reads/writes v No joins supported as RDBMS v Supports Tables with column family, rows and columns v All columns belong to column family v Have table cells with intersection of row and column coordinates and are versioned {row, column, version} v You can run get command to select row, put to insert or update a row, scan to do a loop for multiple rows and Delete to delete record By: Milind Zodge v If you have millions or billions of rows, then HBase is a good candidate v You do not need advanced query language SQL and can leave without secondary indexes and typed columns v You have enough hardware available as even HDFS doesn’t do well with anything less than 5 Data Node so you will at least need 5 nodes cluster v If your application has a variable schema where each row is slightly different v If you data I stored in collections v If you need key based access to data when storing or retrieving About HBase When to use HBase