RFR: 8328273: sun/management/jmxremote/bootstrap/RmiRegistrySslTest.java failed with java.rmi.server.ExportException: Port already in use

[Jaikiran Pai]

On Mon, 1 Apr 2024 02:02:40 GMT, Jaikiran Pai <jpai at openjdk.org> wrote:

> Can I please get a review of this change which proposes to fix the test failure reported in https://bugs.openjdk.org/browse/JDK-8328273?
> 
> As noted in that issue, the `sun/management/jmxremote/bootstrap/RmiRegistrySslTest.java` intermittently fails with a port already in use error. The test attempts to find a free port and then uses it during the test. The interesting part is that the test already has a loop of 10 attempts to retry the test if the port wasn't actually free. So for the test to fail, it would then mean that each of the 10 attempts of using a free port failed (which should be extremely rare and should almost never happen). 
> 
> I didn't have an answer for that until today and had it on my TODO to look further. Credit goes to Kevin @kevinjwalls for identifying the issue - turns out this is the exact same issue that Kevin fixed in https://github.com/openjdk/jdk/pull/18470 for a different test. After noticing that fix, I spotted the same typo in the exception message check in this test. That explains why it wasn't retrying at most 10 times. The test was thus immediately failing on first attempt whenever the chosen free port was in use.
> 
> I have run this test with a test repeat of 50 with this change and the test now passes always. Without this change and a test repeat of 50, the test failed 2 times. I've additionally searched for any other similar typos in other tests and haven't found any (I searched for the string "Exception thrown by the agent :").

Thank you Dan for the review. 

Since this is a test-only trivial fix, I'll go ahead with the integration.

-------------

PR Comment: https://git.openjdk.org/jdk/pull/18561#issuecomment-2030898184