I have more than 5000 hive cluster tasks, calculated daily,
At the same time, I will support the temporary cluster the data extraction, namely, the query non-routing tasks, such as
Cluster is the total amount of data has 10 p

I hope the data don't copy, just a, but routine tasks and non-routing task (temporary) computing resources can be separated, not interfere with each other, can you tell me how to do?