-
Bug
-
Resolution: Unresolved
-
Critical
-
Narayana-LRA-0.0.10.Final
-
None
Probably after the WildFly PR https://github.com/wildfly/wildfly/pull/18499 the crash tests fail. In the PR the recovery scan is triggered during the startup of the Service.
I attached the failsafe reports of the failing test (failsafe-reports-arq.zip) and the logs from an older build (executed before the WildFly PR got merged) which worked (WORKING-failsafe-reports-arq.zip).
Tests run: 2, Failures: 1, Errors: 1, Skipped: 0, Time elapsed: 61.133 s <<< FAILURE! - in io.narayana.lra.arquillian.LRACoordinatorRecoveryIT
lraCoordinatorRecoveryTwoLRAs(io.narayana.lra.arquillian.LRACoordinatorRecoveryIT) Time elapsed: 37.139 s <<< ERROR!
jakarta.ws.rs.WebApplicationException: HTTP 412 Precondition Failed
at io.narayana.lra.arquillian.LRACoordinatorRecoveryIT.lraCoordinatorRecoveryTwoLRAs(LRACoordinatorRecoveryIT.java:250)
lraCoordinatorShortTimeoutLRA(io.narayana.lra.arquillian.LRACoordinatorRecoveryIT) Time elapsed: 22.623 s <<< FAILURE!
java.lang.AssertionError: A new LRA should have been added to the object store before the JVM was halted.
at io.narayana.lra.arquillian.LRACoordinatorRecoveryIT.lraCoordinatorShortTimeoutLRA(LRACoordinatorRecoveryIT.java:174)
- depends on
-
JBTM-3971 Update the state of timed out LRAs when recovering them after a crash
- Open
- is triggered by
-
WFLY-20026 Delayed transaction visibility in LRA coordinator after server restart
- Closed