As part of this topic, we will see
- Architecture of HBase
- Setting up HBase
- Plain Vanilla
- Hortonworks Ambari
- Post setup demo
Setup videos are picked up from administration videos. It is just added for informational purposes for developers.
Let us see the overview of HBase.
- HBase is a NoSQL Datastore which uses HDFS as storage and can be used for applications that are operational in nature.
- Recommendation engines
- Endorsement engines
- It might not be good solution where transactions are important.
- It addresses scalability and can complement both operational as well as decision support applications
HBase Architecture – Details
HBase have several components
- HBase APIs – Programming APIs to perform CRUD operations
- Zookeeper – for High Availability
- HMaster – typically 3
- HFile – Data will be stored in HFiles which will be persistently saved in HDFS
- Mem Store – Data will be first stored into memory
- Write Ahead log – For recovery purpose, all insert/update operations will be logged in WAL
- HDFS – Data will be persistently stored in HDFS
Setup of HBase on Plain Vanilla Hadoop
Setup of HBase using Ambari