Open better629 opened 2 years ago
Do you add other dependency? This error seem like dependency conflict. Also please provide your config.
no other dependency, and the config is
source {
kafkaStream {
topics = "test_topic"
consumer.bootstrap.servers = "xxx1:9092,xxx2:9092,xxx3:9092"
consumer.group.id = "seatunnel_group"
}
}
transform {
json {
source_field = "raw_message"
result_table_name = res_table
}
}
sink {
Hive {
source_table_name = "res_table"
result_table_name = "xxx.seatunnel_test"
save_mode = "append"
sink_columns = "id,key"
}
}
@better629 Can you read write data into hive success with your spark environment? It seems this caused by your spark job has not loaded the jar libfb303-0.9.3.jar
, and this jar should in your ${spark_home}/jars
@ruanwenjun the jar exists, it's
xxx]# find ./ -name "*fb303*"
./jars/libfb303-0.9.3.jar
@better629 i have got the same problem,did you solve it?
@someorz I have tried version seatunnel-1.5.7, and it works. Maybe you can have a try. But the above problem not solved.
@better629 modify root pom.xml maven-shade-plugin,exclude org.apache.thrift:libthrift
@someorz how to find this?
@better629 analyza dependencies
Hive actually uses libthrift-0.9.3
instead of libthrift-0.9.0
, but due to dependency conflicts, the packaged jar is libthrift-0.9.0
.
You can try to exclude the libthrift-0.9.0
dependency in the seatunnel-connectors-spark-dist
module.
Search before asking
Description
Have any one have load data from kakfa to hive successed? Can u share a example?
Seatunnel 2.1.0(compiled) Hive 2.5.7 Spark 2.4.0 Hadoop 2.8.0
Although I can load data from kafka to hive table sometimes: Hive Table:db_test.db/test_table/user_10.db/part-00000-e7732d9e-e4ff-4499-8c6a-60216ba51fdb-c000.snappy.parquet
And has a error:
But I can table test_table not in
show tables
and the data in parquet file is only part of the produce messages.Usage Scenario
-
Related issues
-
Are you willing to submit a PR?
Code of Conduct