1 Big D, Hdfs, Map R TEST

Without replication, a 500 MB file in HDFS occupies 500 MB.

The replication factor is configurable per file, with a default value of 3.

The client writes blocks to a datanode, which then propagates the write command to other datanodes.

It's impossible for two replicas of the same block to reside on the same node.

None of the provided answers are correct.

HDFS is scalable because you can increase cluster capacity by adding more nodes.

HDFS files are typically large, created once, not modified, and frequently read.

The failure of a namenode only causes a service interruption of about one minute before another namenode takes over.

MapReduce involves writing to disk and moving data between machines between the map and reduce phases.

Add more machines to the cluster, leveraging the existing infrastructure.

GFS (Google File System).

A set of Hadoop ecosystem applications with interoperable versions ready to use.

Google.

Technologies to store, move, process, and analyze immense amounts of heterogeneous data.

Unstructured data generated by people.

Volume, velocity, and variety.

Machines of normal power and cost connected to form a more powerful cluster.

The most effective way involves Big Data technologies.