Q#$01.$ Consider a distributed storage system utilizing XOR$-$based erasure coding for fault tolerance. In this system, data is divided into blocks, and a parity block is generated using XOR operations. $(6+4+5$ Marks$)$

$4.$ Given a scenario where data blocks $A,B,$ and C are encoded into parity block P using XOR operations $),$ demonstrate how the system can reconstruct if data block C fails of the data blocks $($A$,$ B$,$ or C $)$ using the available data and parity block.

B$.$ Discuss the advantages and limitations of XOR$-$based erasure coding compared to replication scheme.

C$.$ What's the Catch w$.$r$.$t the given scenario.

Q#$02.$ Discuss the key limitations of Hadoop $1\mathrm{.}0$ compared to Hadoop $2\mathrm{.}0.$

$(5$ Marks$)$

Q#$03.$ Discuss the relation behind storing data on mapper's local disk rather than on HDFS$.$ Additionally, explain under what circumstances the Mapper output may be written to HDFS$.$

$(5$ Marks$)$

Q#$04.$ Let's consider given scenario with the Hadoop framework operating with four mappers: Mapper $1,$ Mapper $2,$ Mapper $3,$ and Mapper $4.$ Each mapper receives one record of a $log$ file as input, with the key representing a text string in the format "file name $+$ line number." The mapper processes each log record to generate key$-$value pairs, with a placeholder value of $\text{'}1\text{'}$ for simplicity. The resulting output from the mappers is as follows:

A$.$ Run combiners against the given Mappers also write down output of each combiner.

B$.$ How the partitioner allocates the data from the combiners to the reducers and what will be the input to the reducers?

C$.$ How the partitioner allocates the data if there were no combiners involved, what will be the input to the reducers?

D$.$ Write down output of reducer.