融合的HDFS Sink连接器：Kafka主题以纯字符串格式转换为Parquet格式的HDFS失败，并且Avro模式必须是记录错误

Question

我已经在虚拟机中安装了docker confluentinc / cp-kafka-connect：4.0.0映像。 我对获取kafka主题感兴趣，该主题是镶木地板格式的hdfs中的纯文本数据（字符串格式）。

sudo docker run -d \
    --name=kafka-connect \
    --net=host \
    -e CONNECT_BOOTSTRAP_SERVERS=<kafka-server-host>:9092 \
    -e CONNECT_REST_PORT=8082 \
    -e CONNECT_GROUP_ID="connect-kafkac1" \
    -e CONNECT_CONFIG_STORAGE_TOPIC="connect-kafkac1-config" \
    -e CONNECT_OFFSET_STORAGE_TOPIC="connect-kafkac1-offsets" \
    -e CONNECT_STATUS_STORAGE_TOPIC="connect-kafkac1-status" \
    -e CONNECT_KEY_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
    -e CONNECT_VALUE_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
    -e CONNECT_INTERNAL_KEY_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
    -e CONNECT_INTERNAL_VALUE_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
    -e CONNECT_REST_ADVERTISED_HOST_NAME="localhost" \
    -e CONNECT_LOG4J_LOGGERS=org.reflections=ERROR \
    -e CONNECT_PLUGIN_PATH=/usr/share/java \
    confluentinc/cp-kafka-connect:4.0.0

我已完成以下配置。

/etc/kafka/connect-standalone.properties

bootstrap.servers=kafkaclusteraddress:9092
key.converter=org.apache.kafka.connect.storage.StringConverter
value.converter=org.apache.kafka.connect.storage.StringConverter
key.converter.schemas.enable=false
value.converter.schemas.enable=false
internal.key.converter=org.apache.kafka.connect.json.JsonConverter
internal.value.converter=org.apache.kafka.connect.json.JsonConverter
internal.key.converter.schemas.enable=false
internal.value.converter.schemas.enable=false

/etc/kafka-connect-hdfs/quickstart-hdfs.properties

name=hdfs-sink
connector.class=io.confluent.connect.hdfs.HdfsSinkConnector
tasks.max=1
topics=<topicname>
topics.dir=<hdfs-topic-dir>
logs.dir=<hdfs-logs-dir>
hdfs.url=<hdfs-url>:8020
format.class=io.confluent.connect.hdfs.parquet.ParquetFormat
flush.size=3
hadoop.conf.dir=/etc/hadoop/conf
hadoop.home=/usr/bin/hadoop

在独立模式下运行配置后收到的错误

# connect-standalone /etc/kafka/connect-standalone.properties /etc/kafka-connect-hdfs/quickstart-hdfs.properties

----------------------------------------

[2018-02-13 19:11:09,542] ERROR WorkerSinkTask{id=hdfs-sink-0} Task threw an uncaught and unrecoverable exception. Task is being killed and will not recover until manually restarted. (org.apache.kafka.connect.runtime.WorkerSinkTask)
java.lang.IllegalArgumentException: Avro schema must be a record.
        at org.apache.parquet.avro.AvroSchemaConverter.convert(AvroSchemaConverter.java:113)
        at org.apache.parquet.avro.AvroParquetWriter.writeSupport(AvroParquetWriter.java:144)
        at org.apache.parquet.avro.AvroParquetWriter.<init>(AvroParquetWriter.java:106)
        at io.confluent.connect.hdfs.parquet.ParquetRecordWriterProvider$1.write(ParquetRecordWriterProvider.java:68)
        at io.confluent.connect.hdfs.TopicPartitionWriter.writeRecord(TopicPartitionWriter.java:635)
        at io.confluent.connect.hdfs.TopicPartitionWriter.write(TopicPartitionWriter.java:379)
        at io.confluent.connect.hdfs.DataWriter.write(DataWriter.java:374)
        at io.confluent.connect.hdfs.HdfsSinkTask.put(HdfsSinkTask.java:101)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.deliverMessages(WorkerSinkTask.java:495)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.poll(WorkerSinkTask.java:288)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.iteration(WorkerSinkTask.java:198)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.execute(WorkerSinkTask.java:166)
        at org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:170)
        at org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:214)
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
        at java.util.concurrent.FutureTask.run(FutureTask.java:266)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
        at java.lang.Thread.run(Thread.java:745)
[2018-02-13 19:11:09,545] ERROR WorkerSinkTask{id=hdfs-sink-0} Task threw an uncaught and unrecoverable exception (org.apache.kafka.connect.runtime.WorkerTask)
org.apache.kafka.connect.errors.ConnectException: Exiting WorkerSinkTask due to unrecoverable exception.
        at org.apache.kafka.connect.runtime.WorkerSinkTask.deliverMessages(WorkerSinkTask.java:517)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.poll(WorkerSinkTask.java:288)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.iteration(WorkerSinkTask.java:198)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.execute(WorkerSinkTask.java:166)
        at org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:170)
        at org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:214)
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
        at java.util.concurrent.FutureTask.run(FutureTask.java:266)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
        at java.lang.Thread.run(Thread.java:745)
Caused by: java.lang.IllegalArgumentException: Avro schema must be a record.
        at org.apache.parquet.avro.AvroSchemaConverter.convert(AvroSchemaConverter.java:113)
        at org.apache.parquet.avro.AvroParquetWriter.writeSupport(AvroParquetWriter.java:144)
        at org.apache.parquet.avro.AvroParquetWriter.<init>(AvroParquetWriter.java:106)
        at io.confluent.connect.hdfs.parquet.ParquetRecordWriterProvider$1.write(ParquetRecordWriterProvider.java:68)
        at io.confluent.connect.hdfs.TopicPartitionWriter.writeRecord(TopicPartitionWriter.java:635)
        at io.confluent.connect.hdfs.TopicPartitionWriter.write(TopicPartitionWriter.java:379)
        at io.confluent.connect.hdfs.DataWriter.write(DataWriter.java:374)
        at io.confluent.connect.hdfs.HdfsSinkTask.put(HdfsSinkTask.java:101)
        at org.apache.kafka.connect.runtime.WorkerSinkTask.deliverMessages(WorkerSinkTask.java:495)
        ... 10 more
[2018-02-13 19:11:09,545] ERROR WorkerSinkTask{id=hdfs-sink-0} Task is being killed and will not recover until manually restarted (org.apache.kafka.connect.runtime.WorkerTask)

Answer 1

如果您查看堆栈跟踪，它期望的是Avro数据，而不是字符串。

IllegalArgumentException: Avro schema must be a record.
        at org.apache.parquet.avro.AvroSchemaConverter.convert(AvroSchemaConverter.java:113)
        at org.apache.parquet.avro.AvroParquetWriter.writeSupport(AvroParquetWriter.java:144)
        at org.apache.parquet.avro.AvroParquetWriter.<init>(AvroParquetWriter.java:106)

我没有亲自测试JSON到Parquet，但是作为列式数据类型， ParquetFormat要求您具有模式。 由于您具有schemas.enable=false ，因此不会发生，因此它无法使用纯字符串，整数，布尔值等以及其他“ schemas.enable=false ” Connect API类型。

这些是需要更改的设置。

key.converter=org.apache.kafka.connect.storage.StringConverter
value.converter=org.apache.kafka.connect.storage.StringConverter
key.converter.schemas.enable=false
value.converter.schemas.enable=false

为了将JSONConverter设置为具有schemas.enable=true ，那么您必须生成看起来像{"schema": {...}, "payload": {...}} ，其中schema字段包含payload字段中对象的类型定义

否则，如果您可以控制生产者代码，则应改为发送Avro数据，这需要使用Schema Registry。

---

我要提到的是，由于您只有一个字符串记录，因此Parquet与HDFS上的原始文本文件相比不会有太多好处，因为总会只有一列。