What happens sqoop fails in between proceed of data

Question

What happens when Sqoop import job fails while importing data into RDBMS-HDFS and vice-versa?

Answer 1

Sqoop can export data from HDFS into an RDBMS using parallel data transfer tasks. Each task will open a connection to the database, insert into the database via transactions, and commit periodically. This means that before the entire export job is complete, partial data will be available in the database.

If an export map task fails even after multiple retries, the entire job will fail. The reasons for task failures could include network connectivity issues, database integrity constraints, malformed records on HDFS, cluster capacity issues etc. In such a failure case, the already committed data will still be available in the database.