[FLINK-15651] JarHandlerTest.testRunJar fails on travis

• Type: Bug
• Status: Closed
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 1.10.0
• Fix Version/s: 1.10.0
• Component/s: Runtime / REST
• Labels: pull-request-available
• Assignee: Chesnay Schepler
• Reporter: Zhu Zhu
• Created: 19/Jan/20 03:30
• Updated: 03/Feb/20 14:41
• Resolved: 29/Jan/20 08:37

Description

The test JarHandlerTest.testRunJar fails on travis with error below:

09:48:58.828 [ERROR] Tests run: 2, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 2.061 s <<< FAILURE! - in org.apache.flink.runtime.webmonitor.handlers.JarHandlerTest
09:48:58.829 [ERROR] testRunJar(org.apache.flink.runtime.webmonitor.handlers.JarHandlerTest)  Time elapsed: 0.2 s  <<< FAILURE!
java.lang.AssertionError: 

Expected: a string containing "ProgramInvocationException"
     but: was "[Service temporarily unavailable due to an ongoing leader election. Please refresh.]"
	at org.apache.flink.runtime.webmonitor.handlers.JarHandlerTest.runTest(JarHandlerTest.java:125)
	at org.apache.flink.runtime.webmonitor.handlers.JarHandlerTest.testRunJar(JarHandlerTest.java:74)

Seems the cause is the client sends request before the leader election is done.

https://api.travis-ci.com/v3/job/276751147/log.txt

Issue Links

• is related to: FLINK-15504 Flink suppresses stdout/stderr during JobGraph generation from JAR
• links to: GitHub Pull Request #10933