Single Point of Failure in Hadoop (Godown Analogy Part 2)
Name Node and Job tracker is called Single point of failure But how? let me explain in very layman language
Name Node is Manager and Job Tracker is assistant to Name Node, Now we know that all the record of a farmer is recorded by Godown Manager and if Godown manager and its report are somehow missing then everything is gone, there is no way to get the record, means Farmer loses there crops.
Similarly, if the job tracker went down means no information will reach to name node and then again in this case also farmer cannot able to access their crops.
This analogy is also called the Master-Servant analogy. Let me explain this concept also, In big data terminology Name Node and Job Tracker is acting as Master and Task Tracker and Data Node are acting as Servant, Basically, if Masters gone than a servant will not listen because there are no masters to guide them and if the server goes down there is ability in Hadoop to get another servant, as a real life example if you are owning some shop and if you are gone shop will shut down but if you losing your servant you will find another servant for your shop, likewise same with Hadoop cluster.
There is no way to overcome this and this is the major disadvantage of the Hadoop cluster.
