I have a requirement to evaluate a Big Data solution based on SolR but I have some questions unanswered after my search. The solution need to provide access and search capabilities to structural data that reside inside HDFS.
So in my understanding solr can index these data and offer access through an API interface. My questions are these:
Indexing Data in HDFS copies them to a solr-only separate area or does it index them directly on HDFS (and there is no need for replication)?
If it is the first case (solr copies data in separate location) where are the Solr indexed data live? Are they stored on HDFS or are they stored in local file system?
If we have the second case that the data are indexed in HDFS directly does this means that I cannot compress my data with a format like parquet or ORC and index them directly?