Kafka

Kafka source connector

Description

To consume data from Kafka , supported Kafka version >= 0.10.0 .

提示

Engine Supported and plugin name

• Spark: KafkaStream
• Flink: Kafka

Options

Spark

name	type	required	default value
topics	string	yes	-
consumer.group.id	string	yes	-
consumer.bootstrap.servers	string	yes	-
consumer.*	string	no	-
common-options	string	yes	-

Flink

name	type	required	default value
topics	string	yes	-
consumer.group.id	string	yes	-
consumer.bootstrap.servers	string	yes	-
schema	string	yes	-
format.type	string	yes	-
format.*	string	no	-
consumer.*	string	no	-
rowtime.field	string	no	-
watermark	long	no	-
offset.reset	string	no	-
common-options	string	no	-

topics [string]

Kafka topic name. If there are multiple topics, use , to split, for example: "tpc1,tpc2"

consumer.group.id [string]

Kafka consumer group id, used to distinguish different consumer groups

consumer.bootstrap.servers [string]

Kafka cluster address, separated by ,

Spark

Flink

format.type [string]

Currently supports three formats

• json
• csv
• avro

format.* [string]

The csv format uses this parameter to set the separator and so on. For example, set the column delimiter to \t , format.field-delimiter=\\t

schema [string]

• csv

  • The schema of csv is a string of jsonArray , such as "[{\"field\":\"name\",\"type\":\"string\"},{\"field\":\"age\ ",\"type\":\"int\"}]" .

• json

  • The schema parameter of json is to provide a json string of the original data, and the schema can be automatically generated, but the original data with the most complete content needs to be provided, otherwise the fields will be lost.

• avro

  • The schema parameter of avro is to provide a standard avro schema JSON string , such as {\"name\":\"test\",\"type\":\"record\",\"fields\":[{ \"name\":\"name\",\"type\":\"string\"},{\"name\":\"age\",\"type\":\"long\"} ,{\"name\":\"addrs\",\"type\":{\"name\":\"addrs\",\"type\":\"record\",\"fields\" :[{\"name\":\"province\",\"type\":\"string\"},{\"name\":\"city\",\"type\":\"string \"}]}}]}

• To learn more about how the Avro Schema JSON string should be defined, please refer to: https://avro.apache.org/docs/current/spec.html

rowtime.field [string]

Extract timestamp using current configuration field for flink event time watermark

watermark [long]

Sets a built-in watermark strategy for rowtime.field attributes which are out-of-order by a bounded time interval. Emits watermarks which are the maximum observed timestamp minus the specified delay.

offset.reset [string]

The consumer's initial offset is only valid for new consumers. There are three modes

• latest
  • Start consumption from the latest offset
• earliest
  • Start consumption from the earliest offset
• specific
  • Start consumption from the specified offset , and specify the start offset of each partition at this time. The setting method is through offset.reset.specific="{0:111,1:123}"

consumer.* [string]

In addition to the above necessary parameters that must be specified by the Kafka consumer client, users can also specify multiple consumer client non-mandatory parameters, covering all consumer parameters specified in the official Kafka document.

The way to specify parameters is to add the prefix consumer. to the original parameter name. For example, the way to specify auto.offset.reset is: consumer.auto.offset.reset = latest . If these non-essential parameters are not specified, they will use the default values given in the official Kafka documentation.

common options [string]

Source plugin common parameters, please refer to Source Plugin for details

Examples

Spark

kafkaStream {
    topics = "seatunnel"
    consumer.bootstrap.servers = "localhost:9092"
    consumer.group.id = "seatunnel_group"
}

Flink

KafkaTableStream {
    consumer.bootstrap.servers = "127.0.0.1:9092"
    consumer.group.id = "seatunnel5"
    topics = test
    result_table_name = test
    format.type = csv
    schema = "[{\"field\":\"name\",\"type\":\"string\"},{\"field\":\"age\",\"type\":\"int\"}]"
    format.field-delimiter = ";"
    format.allow-comments = "true"
    format.ignore-parse-errors = "true"
}