Map Reduce from hadoop ecosystem has gone through lack of popularity considering Spark being preferred for its faster speed and other advantages. For newer projects, Spark is preferred over map reduce. What about other component from hadoop ecosystem like Hbase etc. Is hive the most used component from hadoop ecosystem these days?
I think the answer is Hive. If there is one technology from Hadoop ecosystem which is such frequent in majority of Big Data requirements it is Hive. This makes me wonder is it all for ad hoc analysis. So much of ad hoc analysis? as that is what hive is used for.