[KAFKA-6728] Kafka Connect Header Null Pointer Exception

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 1.1.0
• Fix Version/s: 1.1.1, 2.0.0
• Component/s: connect
• Environment: Linux Mint
• Assignee: Randall Hauch
• Reporter: Philippe Hong
• Created: 30/Mar/18 14:50
• Updated: 25/Apr/18 09:41
• Resolved: 03/Apr/18 15:48

Description

I am trying to use the newly released Kafka Connect that supports headers by using the standalone connector to write to a text file (so in this case I am only using the sink component)

I am sadly greeted by a NullPointerException :

ERROR WorkerSinkTask{id=local-file-sink-0} Task threw an uncaught and unrecoverable exception (org.apache.kafka.connect.runtime.WorkerTask:172)
java.lang.NullPointerException
    at org.apache.kafka.connect.runtime.WorkerSinkTask.convertHeadersFor(WorkerSinkTask.java:501)
    at org.apache.kafka.connect.runtime.WorkerSinkTask.convertMessages(WorkerSinkTask.java:469)
    at org.apache.kafka.connect.runtime.WorkerSinkTask.poll(WorkerSinkTask.java:301)
    at org.apache.kafka.connect.runtime.WorkerSinkTask.iteration(WorkerSinkTask.java:205)
    at org.apache.kafka.connect.runtime.WorkerSinkTask.execute(WorkerSinkTask.java:173)
    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$RunnableAdapter.call(Executors.java:511)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)

I launched zookeeper and kafka 1.1.0 locally and sent a ProducerRecord[String, Array[Byte]] using a KafkaProducer[String, Array[Byte]] with a header that has a key and value.

I can read the record with a console consumer as well as using a KafkaConsumer (where in this case I can see the content of the header of the record I sent previously) so no problem here.

I only made two changes to the kafka configuration:
    - I used the StringConverter for the key and the ByteArrayConverter for the value.
    - I also changed the topic where the sink would connect to.

If I forgot something please tell me so as it is the first time I am creating an issue on Jira.

Issue Links

• is related to: KAFKA-6740 Plugins class' newConverter and newHeaderConverter methods are unclear
• links to: GitHub Pull Request #4800
• links to: GitHub Pull Request #4815