Skip to content
Zhou Yang edited this page Oct 27, 2021 · 4 revisions

Supported data sources

Reader

JDBCReader

JDBCReaders use universal implementations, are different in statement properties.

ClickHouse
clickhouse

MariaDB
mariadb

MySQL
mysql

Oracle
oracle

PostgreSQL
postgresql

Writer

ClickHouse
clickhouse, ClickHouse use MergeTree() as default engine to create table. Default write mode is PREPARED_BATCH mode.

Doris
Doris do not create table.
doris, Doris JDBC support STATEMENT_VALUES mode only.
dorisOverHttp, Doris HTTP use http connector, support JSON load.

MySQL
mysql

PostgreSQL
postgresql, PostgreSQL support COPY FROM STDIN as the default write mode.

Connector

Connector is the component to connect to the data source.

JDBC Connector

hikari is a JDBC Connector, which holding connection to data source.

HTTP Connector

http connector does not hold a connection to data source, but the DataConnection. Each data source with http connector can implement in their ways.

Reporter

Reporter is the component to report message between read batch data and write to sink.

Channel

Channel is used to transfer batch data among reader and writers. Reader push data to channel and, channel notify Writer to write.

Clone this wiki locally