采用Flink SQL 读取kafka 数据，数据格式是avro('value.format' = 'avro-confluent')，任务启动报错了ValidationException: Could not find any factory for identifier 'avro-confluent'
二、错误Caused by: org.apache.flink.table.api.ValidationException: Could not find any factory for identifier 'avro-confluent' that implements 'org.apache.flink.table.factories.DeserializationFormatFactory' in the classpath. Available factory identifiers are: canal-json csv debezium-json json maxwell-json raw at org.apache.flink.table.factories.FactoryUtil.discoverFactory(FactoryUtil.java:319) at org.apache.flink.table.factories.FactoryUtil$TableFactoryHelper.discoverOptionalFormatFactory(FactoryUtil.java:751) at org.apache.flink.table.factories.FactoryUtil$TableFactoryHelper.discoverOptionalDecodingFormat(FactoryUtil.java:649) at org.apache.flink.table.factories.FactoryUtil$TableFactoryHelper.discoverDecodingFormat(FactoryUtil.java:633) at org.apache.flink.streaming.connectors.kafka.table.KafkaDynamicTableFactory.lambda$getValueDecodingFormat$2(KafkaDynamicTableFactory.java:279) at java.util.Optional.orElseGet(Optional.java:267) at org.apache.flink.streaming.connectors.kafka.table.KafkaDynamicTableFactory.getValueDecodingFormat(KafkaDynamicTableFactory.java:277) at org.apache.flink.streaming.connectors.kafka.table.KafkaDynamicTableFactory.createDynamicTableSource(KafkaDynamicTableFactory.java:142) at org.apache.flink.table.factories.FactoryUtil.createTableSource(FactoryUtil.java:134) ... 40 more三、解决
1、avro-confluent 插件注册
2、 发现打包的jar 里面没有meta-INF.services，Flink 加载 table factory 使用的时SPI机制，默认不打包进去的
3、需添加参数
