Using AdoptOpenJDK for Microsoft Integration Runtime

As you may know, Oracle bought Java and has changed the license to no longer be free for commercial use.

You may not want to deal with the license when installing Java which is often necessary for various applications.

Microsoft Azure Integration Runtime for Data Factory is one such application that requires the Java Runtime Environment (JRE). (Java is not required for all IR functionality but is needed for using formats such as Parquet)

While the prerequisites recommend AdoptOpenJDK, they leave out a crucial required step regarding the environment variable.

When you’re installing AdoptOpenJDK, you get this option:

JAVA_HOME env variable setting at installatio…but it still requires more config

You might still be getting an error like this:

Error message in Azure Data Factory about missing Java Runtime

Failure happened on ‘Sink’ side. ErrorCode=UserErrorJreNotFound,’Type=Microsoft.DataTransfer.Common.Shared.HybridDeliveryException,Message=Java Runtime Environment could not be found

But you just installed it, right?!

AdoptOpenJDK installs the whole JDK which includes developer tools, as well as the JRE.

jre under the jdk installation

As pointed out in this ADF troubleshooting article, Azure Data Factory Integration Runtime is expecting the environment variable to point right to the JRE folder, under which there are bin\server folders, with a jvm.dll file in it.

To resolve the issue:

Change the environment variable called JAVA_HOME to end with \jre:

add \jre to the current environment variable JAVA_HOME set by the JDK

Then restart the Integration Runtime service. Or restart the whole machine.

Service to restart when you’re ready

Then you should be all set!

able to use parquet and avro formats from your self hosted integration runtimes.