Migrating data
Also available as:
PDF

Import RDBMS data to HDFS

You create a single Sqoop import command that imports data from diverse data sources, such as a relational database, into HDFS.

You enter the Sqoop import command on the command line of your cluster to import data from a data source into HDFS. In HDFS, you can perform ETL on the data, move the data into Hive, and query the data. The import command needs to include the database URI, database name, and connection protocol, such as jdbc:mysql: and the data to import. Optionally, the command can include parallel processing directives for performant data transfer, the HDFS destination directory for imported data, data delimiters, and other information. The default directory is used if you do not specify another location. Fields are comma-delimited and rows are line-delimited.You can test the import statement before actually executing it.
  • Apache Sqoop is installed and configured.
  1. Create an import command that specifies the Sqoop connection to the data source you want to import.
    • If you want to enter a password for the data source on the command line, use the -P option in the connection string.
    • If you want to specify a file where the password is stored, use the --password-file option.
    Password on command line:
    sqoop import --connect jdbc:mysql://db.foo.com/bar \
    <data to import> \
    --username <username> \
    -P
    Specify password file:
    sqoop import --connect jdbc:mysql://db.foo.com/bar \
    --table EMPLOYEES \
    --username <username> \
    --password-file ${user.home}/.password
  2. Specify the data to import in the command.
    • Import an entire table.
    • Import a subset of the columns.
    • Import data using a free-form query.
    Entire table:
    sqoop import \
    --connect jdbc:mysql://db.foo.com/bar \
    --table EMPLOYEES
    Subset of columns:
    sqoop import \
    --connect jdbc:mysql://db.foo.com/bar \
    --table EMPLOYEES \
    --columns "employee_id,first_name,last_name,job_title"
    Free-form query to import the latest data:
    sqoop import \
    --connect jdbc:mysql://db.foo.com/bar \
    --table EMPLOYEES \
    --where "start_date > '2018-01-01'"
  3. Specify the destination of the imported data using the --target-dir option.
    This command appends data imported from the MySQL EMPLOYEES table to the output files in the HDFS target directory using default text file delimiters.
    sqoop import \
    --connect jdbc:mysql://db.foo.com:3600/bar \
    --table EMPLOYEES \
    --where "id > 100000" \
    --target-dir /incremental_dataset \
    --append
    This command splits imported data by column and specifies importing the data into output files in the HDFS target directory.
    sqoop import \
    --connect jdbc:mysql://db.foo.com:3600/bar \
    --query 'SELECT a.*, b.* \
    FROM a JOIN b on (a.id == b.id) \
    WHERE $CONDITIONS' \
    --split-by a.id \
    --target-dir /user/foo/joinresults
    This command executes once and imports data serially using a single map task as specified by the -m 1 options:
    sqoop import \
    --connect jdbc:mysql://db.foo.com:3600/bar \
    --query \
    'SELECT a.*, b.* \
    FROM a \
    JOIN b on (a.id == b.id) \
    WHERE $CONDITIONS' \
    -m 1 \
    --target-dir /user/foo/joinresults
  4. Optionally, specify write parallelism in the import statement to execute a number of map tasks in parallel:
    • Set mappers: If the source table has a primary key, explicitly set the number of mappers using --num-mappers.
    • Split by: If primary keys are not evenly distributed, provide a split key using --split-by
    • Sequential: If you do not have a primary key or split key, import data sequentially using --num-mappers 1 or --autoreset-to-one-mapper in query.
    • Set mappers:
      sqoop import --connect jdbc:mysql://db.foo.com:3306/bar \
      --table EMPLOYEES \
      --num-mappers 8
    • Split by:
      sqoop import --connect jdbc:mysql://db.foo.com:3306/bar \
      --table EMPLOYEES \
      --split-by dept_id
    • Setting mappers evenly splits the primary key range of the source table.
    • Split by evenly splits the data using the split key instead of a primary key.
  5. Optionally, test the import command before execution using the eval option.
    sqoop eval --connect jdbc:mysql://db.foo.com:3306/bar \
    --query "SELECT * FROM employees LIMIT 10"
    The output of the select statement appears.