Closed famod closed 2 weeks ago
/cc @brunobat
The response being null is a new one... I would bet the second exception triggers the OTel error. There is a request, but it fails to be processed and is null when OTel runs... @cescoffier Should this be addresses by the vertx part of the stack?
Not sure I understand the whole picture. The exception means that Vertx cannot write the response because the event loop is shutdown (which only happens when the application terminates).
The request has been processed 'successfully' as there is a http status to be written (that's the write that fails).
For Otel, I'm. It sure what should be done in this case.
If it's shutting down, and only happens on tests, do we even want to create a span? If no response was written, the only meaningful span result would be a 500 error.
FWIW, I just had the same NPE warning on dev mode reload due to java file change. (while testing Quarkus 3.13.3 if that matters)
Edit: Ugh...now multiple times. Btw, I used to be seeing different exceptions on reload: #29646
So after the NPE fix (thanks!), I'm still getting this long "event executor terminated" exception trace.
I don't think it has anything to do with OTEL because it also pops up when I pass -Dquarkus.otel.enabled=false
to the test execution.
I should probably create another issue?
Not sure @famod. The issue relies on the fact that the execution is aborted and you cannot use certain resources anymore.
Describe the bug
Setting
quarkus.http.test-timeout
to a low value so that a rest assured request times out results in a NPE in OTEL code.Expected behavior
No NPE, just a timeout.
Actual behavior
How to Reproduce?
q_otel-assured.zip
./mvnw clean install -Dquarkus.http.test-timeout=2s
Output of
uname -a
orver
No response
Output of
java -version
No response
Quarkus version or git rev
3.13.2
Build tool (ie. output of
mvnw --version
orgradlew --version
)Maven 3.9.8
Additional information
There is another exception that looks like "noise" too: