Hadoop supports three deployment modes, which are briefly (don't want to deviate too far from the crux of this chapter) detailed here:
- Standalone (Local) Mode
- Default mode
- Single Java (JVM) process
- Non-distributed or non-clustered architecture
- The easiest setup of the three
- Mostly used for learning and running examples and for development
- Very useful for debugging purposes
- Pseudo Distributed Mode (single node cluster)
- Single-node installation
- Each daemon runs in its own Java (JVM) process
- Can be used for simulating multi-node installation/support issues in a single node itself (mini cluster setup on single node)
- Much easier to setup than fully distributed Hadoop installation
- Similar to the first mode, it is non-distributed installation
- Fully Distributed Mode (multi-node cluster)
- Fully distributed clustered architecture
- Mostly used for production deployment
- Can be set using...