Although Spark has superseded Hadoop (MapReduce), it does not have its own storage mechanism, thus, you have to pair it with a storage system to work. Often, professionals use Hadoop-based storage like HDFS.
Here is a video tutorial which you can watch to learn more about spark:-