when i am importing the data from oracle to hive , then table is not created in hive but data will be there in hdfs
I'm not sure of the question, but the answer is yes? ;-)
It really depends on how you're importing the data.... which tools you're using.
If you are using an ETL tool, then yes, the schema will be in hive.
If you're just dropping the exported file from the Oracle system, it will sit on HDFS / MapR-FS. It may or may not be readable by Hive. (It won't unless you've dropped the file in to a directory that is also a hive table.)
Hive is really a SQL like language that allows one to store data in files where the directory tree is the partitioned table. In order for Hive to read the data in the file, it has to know the file's schema. Frameworks like Spark can infer the schema based on the file type and some basic knowledge. Hive doesn't have this built in, so you need to know your schema and create a hive table. (Usually an external table.)
Hive can also be used against data stores like MapR-DB and HBase as well as tsv, ORC, RCFiles, and Parquet files.
It sounds like you're dropping files from Oracle into a directory where you want to be able to then use hive to query these tables?
I am having somewhat same issue. I am creating a database and some external tables in hive. Also I have to create production table in same db. I created the database and the tables but once I exit the hive console and re-entered, the database and tables were gone. Also I am unable to access the db and tables I created from nodes other than the one I created them from. Please help..
Not sure why the tables are gone. In hive, if you execute the create table, it should exist.
Have you tried to do this using beeline instead?
We've noticed with MapR 6.0 that when setting roles, the roles are also not set post session...
Actually, I connected via the Hive Shell as user mapr. I was able to create a database and the create a dumb table.
That said... For some reason or another... Apache is pushing beeline and deprecating Hive. So.. you should be able to connect and use beeline to do this.
Let us know if you're still having trouble with Hive and Beeline.
Retrieving data ...