End Vert.x response before propagating mapped exception in VertxBlockingoutput

[pedroigor]

Relates to: #21841

[pedroigor]

@geoand Took more time than expected because I was trying to figure out how to force the error using tests. I failed, if you think a test is a must, do you have an idea how to force the behavior using a test?

[sberyozkin]

I believe we've had a few of similar fixes which can be hard to reproduce with tests

[gsmet]

Let's wait for @stuartwdouglas 's blessing before merging this, please.

[quarkus-bot]

Failing Jobs - Building 8bea56c

Status	Name	Step	Failures	Logs	Raw logs
✖	Gradle Tests - JDK 11 Windows	Build	Failures	Logs	Raw logs

Full information is available in the Build summary check run.

Failures

⚙️ Gradle Tests - JDK 11 Windows #

- Failing: integration-tests/gradle 

📦 integration-tests/gradle

✖ io.quarkus.gradle.devmode.MultiModuleIncludedBuildTest.main line 24 - More details - Source on GitHub

org.awaitility.core.ConditionTimeoutException: Condition with lambda expression in io.quarkus.test.devmode.util.DevModeTestUtils that uses java.util.function.Supplier, java.util.function.Supplierjava.util.concurrent.atomic.AtomicReference, java.util.concurrent.atomic.AtomicReferencejava.lang.String, java.lang.Stringboolean was not fulfilled within 1 minutes.
	at org.awaitility.core.ConditionAwaiter.await(ConditionAwaiter.java:164)
	at org.awaitility.core.CallableCondition.await(CallableCondition.java:78)

[pedroigor]

@sberyozkin @geoand To reproduce this one we need to somehow force an error on the write stream so that the exception is set here

quarkus/extensions/resteasy-classic/resteasy/runtime/src/main/java/io/quarkus/resteasy/runtime/standalone/VertxBlockingOutput.java

Line 34 in 8bea56c

throwable = event;

.

Then, the write method should be called.

We should be able to check by looking if headers were written, or if handlers registered via RoutingContext (e.g.: body, headers end handler, etc) are invoked.

[stuartwdouglas]

I am ok with skipping a test for this one. We do have a couple of similar tests for the read side such as https://github.com/stuartwdouglas/quarkus/blob/04aa95308cc3296ec89871932cfae101daf9d46a/extensions/resteasy-classic/resteasy/deployment/src/test/java/io/quarkus/resteasy/test/IncompletePostTestCase.java#L38

Basically if you want to test this you need to manually send a HTTP request over a socket, then close the socket after a delay without reading, and also write a large amount of data in the test. Even then I think it will be hard to get a test that passes reliably though.

[pedroigor]

@stuartwdouglas Yeah .. I noticed there are some tests there doing something similar you are suggesting, but I failed to adapt it to my use case :(. But I was not sure what I was doing was the right path, now that you suggested I'll try more.