Ok. It seems if I start hive from the same directory path I used to launch tpds script to generate database/input, the database is recognized. If I change the current location, then not.
For example: hadoop@nectar-11:~/rgrandl/hive/hive-testbench$ ./tpcds-setup.sh 20 if I launch hive from here, the DB is recognized. If I do hadoop@nectar-11:~/rgrandl/hive/hive-testbench$ cd sample-queries-tpcds/ and start hive again, then the DB does not exist. Any idea why is that ? On Saturday, June 21, 2014 4:30 PM, Grandl Robert <rgra...@yahoo.com> wrote: Hi guys, I have created a database using this link: https://github.com/cartershanklin/hive-testbench, which basically creates a database: tpcds_bin_partitioned_orc_20.db and input data using TPC-DS benchmark. The database is visible in HDFS under /user/hive/warehouse/tpcds_bin_partitioned_orc_20.db, but I cannot access it in hive command line. hive.metastore.warehouse.dir=/user/hive/warehouse is correctly set, but when try to do use hive.metastore.warehouse.dir=/user/hive/warehouse it says: hive> use tpcds_bin_partitioned_orc_20; FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask. Database does not exist: tpcds_bin_partitioned_orc_20 However, if I create a database in command line with: create database blah, it is accessible both in HDFS and command line for further operations. Can you guys tell me what I am doing wrong ? I have to mention that I have just started to play with Hive, and I am not very familiar with it. Thanks, robert
