G1Joshi Apache HBase
HBase is the Hadoop database. It is a distributed, scalable, big data store. It provides random, real-time read/write access to your Big Data.
When to Use
- Hadoop Ecosystem: Deep integration with HDFS, Hive, Spark.
- Petabyte Scale: Serving billions of rows with low latency.
- Random Access: When you need random R/W on HDFS data (which is usually WORM - Write Once Read Many).
Quick Start
Uses Java API or Shell.
create 'users', 'info', 'data'
put 'users', 'row1', 'info:name', 'Alice'
get 'users', 'row1'
Core Concepts
Column Families
Data is grouped into column families (info:name, info:email). Families are stored physically together.
Region Servers
HBase scales by splitting tables into "Regions" and hosting them on Region Servers.
WAL & MemStore
Writes go to Write-Ahead-Log (Disk) and MemStore (RAM). When MemStore fills, it flushes to HFile (HDFS).
Best Practices (2025)
Do:
- Design Row Keys carefully: Row keys determine sorting and sharding. "Hotspotting" (sequential keys) is the enemy. Use salt or hashing.
- Pre-split Regions: Don't start with 1 region. Pre-split based on your known key distribution.
- Use Phoenix: Apache Phoenix provides a SQL skin over HBase, making it usable like a Relational DB.
Don't:
- Don't use for small data: The overhead of HDFS/ZimeKeeper/HBase is huge. Only for >TB scale.
- Don't scan excessively: Full table scans are MapReduce jobs.