Create a Hologres result table

create table Hologres_sink(
  name varchar,
  age BIGINT,
  birthday BIGINT
) with (
  type='hologres',
  dbname='<yourDbname>',
  tablename='<yourTablename>',
  username='<yourUsername>',
  password='<yourPassword>',
  endpoint='<yourEndpoint>',
  field_delimiter='|' -- This parameter is optional. 
);

• Exactly-once: The system processes data or events only once even multiple faults occur.
• At-least-once: If streaming data or events to be processed are lost, the system transfers the data again from the transmission source. Therefore, the system may process streaming data or events for multiple times. If the first retry succeeds, no further retries are required.

• If no primary keys have been configured in the Hologres physical table, the Hologres streaming sink uses the at-least-once semantics.
• If primary keys have been configured in the Hologres physical table, the Hologres streaming sink uses the exactly-once semantics based on the primary keys. If multiple records with the same primary key are written to the table, you must set the mutationType parameter to determine how the result table is updated. This parameter has the following valid values:
  • insertorignore (default value): Hologres keeps the first record and discards the subsequent records.
  • insertorreplace: Hologres completely replaces the existing record with the one that arrives later.
  • insertorupdate: Hologres partially replaces the existing record with the one that arrives later.
  Note

  • If the mutationType parameter is set to insertorupdate or insertorreplace, the system updates data based on the primary key.
  • The number of columns in the result table defined by Blink can be different from the number of columns in the Hologres physical table. Make sure that the value null can be used to fill the missing columns. Otherwise, an error is returned.
• By default, the Hologres streaming sink can import data to only one non-partitioned table. If the sink imports data into the parent table of a partitioned table, data queries fail even if data import succeeds. To enable data to be automatically written to a partitioned table, you can set the partitionRouter parameter to true. Take note of the following points:
  • You must set tablename to the name of the parent table.
  • Blink connectors do not automatically create partitioned tables. We recommend that you create a partitioned table before you import data. Otherwise, the data fails to be imported.