Import SQL Server table to HDFS

Question

I am new to Hadoop and I have few queries regarding importing a table from SQL Server to hdfs using sqoop:

1. How to import whole database into hadoop using sqoop and any other way.
2. How can I import a single table which don't have primary key or is it possible to import SQL Server table which don't have primary key.

Answer 1

1. How to import whole database into hadoop using sqoop and any other way.

   You can use sqoop-import-all-tables to do that. Other way would be to write your own job using DBInputFormat.

2. How can I import a single table which don't have primary key or is it possible to import SQL Server table which don't have primary key.

   You can use --split-by to specify a column other than primary key column.

Answer 2

For the 2nd question-How can I import a single table which don't have primary key or is it possible to import SQL Server table which don't have primary key.

When the table we import does not have a primary key we need to specify in the commands as to how many mappers are to be used. this can be specified by the options "-m" or "--Split-by" in the sqoop import

e.g:

$ bin/sqoop import -connect jdbc:mysql://localhost:3306/db1 -username root -password password --table tableName --target-dir /user/tableName -m 1