flink-runtime-web | Default UI in Flink | SQL Database library

For an unknown reason, setting the classloader.resolve-order property to parent-first as mentioned in the Apache Flink mailing list resolves the issue. I am still baffled as to why it works, as there should be no dependency clashes between the child and parent classloader loading different versions of this dependency (as it is not provided out of the box with the flink-dist I am using).

In the Flink documentation under "Debugging Classloading", there's a section which talks about this parent-child relationship:

In setups where dynamic classloading is involved (plugin components, Flink jobs in session setups), there is a hierarchy of typically two ClassLoaders: (1) Java’s application classloader, which has all classes in the classpath, and (2) the dynamic plugin/user code classloader. for loading classes from the plugin or the user-code jar(s). The dynamic ClassLoader has the application classloader as its parent.

By default, Flink inverts classloading order, meaning it looks into the dynamic classloader first, and only looks into the parent (application classloader) if the class is not part of the dynamically loaded code.

The benefit of inverted classloading is that plugins and jobs can use different library versions than Flink’s core itself, which is very useful when the different versions of the libraries are not compatible. The mechanism helps to avoid the common dependency conflict errors like IllegalAccessError or NoSuchMethodError. Different parts of the code simply have separate copies of the classes (Flink’s core or one of its dependencies can use a different copy than the user code or plugin code). In most cases, this works well and no additional configuration from the user is needed.

I have yet to understand why loading ProducerRecord happens more than once, or what this "different type" in the exception message refers to (greping on the result of -verbose:class yielded only a single path for ProducerRecord).

You are using the wrong Kafka jar. You should depend on org.apache.kafka:kafka-clients instead of org.apache.kafka:kafka_2.xx, which is the Kafka server side jar.

The dependence on kafka/api/OffsetRequest is coming from storm-kafka, which should not be used. It's using an old Kafka client API which is no longer present in Kafka. Use storm-kafka-client instead.

I've run into the same issue and solved it by looking at the network request in chrome when uploading a jar with the web UI.

The request must

• Use multipart upload
• The field name must be jarfile
• The multi part content must include the file Content-Type as well (otherwise you'll get a 500 from Flink complaining about the missing header)

Here is a python script using requests that does the upload

Figured it out:

I needed to import StreamTableEnvironment from a specific package: org.apache.flink.table.api.java.StreamTableEnvironment. My autocompletion was not being properly update in time, which made me think there was no proper method. But it is there.