I am trying to understand these to attributes and how the work.
Please someone that tell me if I am wrong but we must set mapred.min.split.sizein our convenience if we are using HDFS files
But if we are reading from hive to cassandra we should set
To give a little of context, we have a cluster cassandra and we do our queries using hive to cassandra. We are experimenting some OOM problems with java heap and we think we must modify one or both of these attributes.