What is best for combining Hive Jive connections

Question

I am using the Hive JDBC driver to execute a sql query in my HDFS datastore. I am trying to use c3p0 to handle a connection pool. I’m not sure if this is the right approach, since the Hive request can take a lot of time, which means that the connection will not be released back to the pool for a long time, I'm struggling to think about the correct installation number for the maximum number of connections in the c3p0 configuration .

Is there any best practice for joining a hive jdbc connection? c3p0? DBHP?

What about MAX_POOL_SIZE? Should it be more than the usual setup for RDB?

Answer 1

, , , , :) , .

, Hive Hadoop, , . , , , , , . Hadoop first-in-first-out (FIFO), . , Fair scheduler Capacity.

, , .

, . -, , . , . ( - ). -, . , .