I've spent a longer while debugging problems related to "hung" test runs on the FreeBSD VM we have Jenkins on. I no longer think it's our fault, really. To me the JVM (OpenJDK distro) on that system is just buggy, see SOLR-6204, for example. The best explanation I have for hung tests (LUCENE-5786) is that the test thread just dies in the middle of executing Java code, although I cannot replicate that behavior. The test thread just vanishes -- it's not present in stack traces and it definitely doesn't execute finally blocks, which it should have executed to terminate.
Can we have a different operating system to test on? :) Dawid --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
