apache inlong FlinkSQL 如何对接? #2952
-
from @walkhan |
Beta Was this translation helpful? Give feedback.
Replies: 4 comments 5 replies
-
@dockerzhang 目前场景采集端那边要log日志,binlog,以及API都要用到。sort端,用flink web 封装平台dinky运行纯SQL,不写任何java代码。dinky git地址:https://github.com/DataLinkDC/dlink。如果apache inlong比较好用,后期打算将apache inlong集成到dinky中去。 |
Beta Was this translation helpful? Give feedback.
-
1,采集后的数据,InLong 支持直接从Pulsar/Kafka 消费,flinkSQL可以直接对接Pulsar/Kafka即可;这里不一样的地方是,从MQ中出来的消息是InLongMsg协议,需要线进行解码,解码逻辑可以参考https://inlong.apache.org/docs/next/development/inlong_msg |
Beta Was this translation helpful? Give feedback.
-
“dataproxy是否支持正则表达式?”,dataproxy在数据流向中用户不可见,用户不需要对dataproxy做任何操作和配置 |
Beta Was this translation helpful? Give feedback.
-
we are implementing #3658 to support the flink SQL, it will release in the next version. |
Beta Was this translation helpful? Give feedback.
1,采集后的数据,InLong 支持直接从Pulsar/Kafka 消费,flinkSQL可以直接对接Pulsar/Kafka即可;这里不一样的地方是,从MQ中出来的消息是InLongMsg协议,需要线进行解码,解码逻辑可以参考https://inlong.apache.org/docs/next/development/inlong_msg
2,配置数据流时不选择具体流向,选“自主消费”即可
3,sort standalone不依赖flink也可以进行分拣,当前该模块还在完善中,当前可用sort flink进行分拣