Godown analogy helps me to understand Big Data architecture more easily.
I am a farmer and I used to grow wheat and I have very limited storage capacity than what should I do if I grow more than the storage capacity I have?
I will approach the nearest godown means I am a client, whom should I approach Godown manager means the manager is acting as the name node. what should a manager do he will check whether the room is available for storage or not, for this work he will tell his helper to go and check.
In big data term, this helper is known as a job tracker, this helper will go to every room and ask worker from the respective room whether there is space available or not, If we talk in big data term, the room is a data node, and worker in the room who is checking whether there is space or not is acting like task tracker.
If there is space available in any room; worker from the respective room will inform the helper of the manager to convey the message that in this particular room there is the availability of space, now in big data term task tracker is giving a message to the job tracker and this job tracker will convey the message to the name node.
Now this manager will inform me that he has a room available for storage and tell me to keep my wheat in the respective room, I will directly put my wheat in that particular room, Manager will keep the record of my no of a packet which I stored in the godown. In Big data terms, the Client has access to the Data node and clients will put their data in the respective data node while the Name node is keeping the record of metadata.
Below is the schematic approach to understanding
I will talk more on Big data stuff incoming Blog, Maybe I am wrong also but this is how I understood.