Introduction
HBase provides various ways to leverage the potential of MapReduce based on the stack and the architecture you are going to use.
Before we start, let's do a quick revisit to the components, which will be used in MapReduce:
- Record reader
- Mapper
- Combiner
- Practitioner
- Shuffle and sort
- Reduce
- Output format
- Record reader: The core responsibility of a record reader is to analyze the data and then parse the data in key-value. The key is the location in the index and the value is the data that is composed of records.
- Mapper: Mapper executes each key-value pair that we got from the records. The design of the key and values depends on what we are planning to achieve from it. The key is the data we will use to group the values.
- Combiner: Combiner is an alternative localized reducer; the main advantage is the ability to group data during the mapping process. It gathers all the in-between keys that are parsed from the previous process (Mapper) and invokes a custom method to rearrange values...
