HDFS

The Hadoop Distributed File System is the main or, more important, core of the Hadoop Ecosystem.
HDFS is the one, which enables us to store a wide variety of large data sets (eg structured, unstructured and well-structured data).
HDFS creates the output of overhead resources, where we can see the whole HDFS as one unit.
It helps us to store our data at all different nodes and to keep a log file about the database (metadata).
HDFS has two main features, namely NameNode and DataNode.
NameNode is the main node and does not store real data. Contains metadata, such as a log file or as a table of contents. Therefore, it requires less storage and higher computer resources.
On the other hand, all your data is stored in DataNode and therefore requires a lot of storage resources. These DataNodes are hardware hardware (like your laptops and desktops) in distributed space. That’s why, why Hadoop solutions are so expensive.
You are always talking to NameNode while typing. After that, it sends in a request to the client to store and duplicate the data to various DataNode.