Phoenix

Phoenix sink connector

Description

Write Phoenix data through Jdbc connector. Support Batch mode and Streaming mode. The tested Phoenix version is 4.xx and 5.xx On the underlying implementation, through the jdbc driver of Phoenix, execute the upsert statement to write data to HBase. Two ways of connecting Phoenix with Java JDBC. One is to connect to zookeeper through JDBC, and the other is to connect to queryserver through JDBC thin client.

Tips: By default, the (thin) driver jar is used. If you want to use the (thick) driver or other versions of Phoenix (thin) driver, you need to recompile the jdbc connector module

Tips: Not support exactly-once semantics (XA transaction is not yet supported in Phoenix).

Key features

• exactly-once
• schema projection

Options

driver [string]

if you use phoenix (thick) driver the value is org.apache.phoenix.jdbc.PhoenixDriver or you use (thin) driver the value is org.apache.phoenix.queryserver.client.Driver

url [string]

if you use phoenix (thick) driver the value is jdbc:phoenix:localhost:2182/hbase or you use (thin) driver the value is jdbc:phoenix:thin:url=http://localhost:8765;serialization=PROTOBUF

Example

use thick client drive

    Jdbc {
        driver = org.apache.phoenix.jdbc.PhoenixDriver
        url = "jdbc:phoenix:localhost:2182/hbase"
        query = "upsert into test.sink(age, name) values(?, ?)"
    }

use thin client drive

    Jdbc {
        driver = org.apache.phoenix.queryserver.client.Driver
        url = "jdbc:phoenix:thin:url=http://spark_e2e_phoenix_sink:8765;serialization=PROTOBUF"
        query = "upsert into test.sink(age, name) values(?, ?)"
    }