There is not much detail in your use case provided.
What is the source of the data?
Do you plan on manipulating and/or filtering the data in anyway before writing to HDFS?
Is this a one time move or a sustained dataflow?
Bottom line is the NiFi putHDFS processor can be used to write data to HDFS. Before NiFi can do this, it must ingest the data from some source. As part of this ingestion, the data is written to the NiFi content repository. A NiFi FlowFile(s) are generated. NiFi will also track lineage information about every FlowFile as it traverses any number of processor between ingestion and termination in NiFi.