5. HBase Physical Architecture

Figure 5.1 shows the deployment view for HBase cluster:
HBase is the master-slave cluster on top of HDFS. The classic deployment is as follows:
➢** Master node:** one HMaster and one NameNode running on a machine as the master node.
➢ Slave node: Each node is running one HRegionServer and one DataNode. And each node report status to the master node and Zookeeper.
➢** Zookeeper:** HBase is shipped with ensemble Zookeeper, but for large clusters, using existing Zookeeper is better. Zookeeper is crucial, the HMaster and HRegionServers will register on Zookeeper.
➢ Client: There can be many clients to access HRegionServer, like Java Client, Shell Client, Thrift Client and Avro Client

Continue reading

4. Process Architecture

4.1 HBase Write Path

The client doesn’t write data directly into HFile on HDFS. Firstly it writes data to WAL(Write Ahead Log), and Secondly, writes to MemStore shared by a HStore in memory.

Continue reading

1. Overview

Apache HBase is an open source column-oriented database. It is often described as a sparse, consistent, distributed, multi-dimensional sorted map. HBase is modeled after Google’s “Bigtable: A distributed Storage System for Structured Data”, which can host very large tables with billions of rows, X millions of columns.

Continue reading