It's not clear why this test fails for some Windows users. It likely has to do with using the Java Chronicle on that platform. The OpenHFT libraries rely on Unsafe to gain their speed for some functions and, in all honesty, I'm not sure how well-supported Java Chronicle is on Windows platforms.
It would be good to have a GitHub issue detailing this failure and including important details about OS, hardware, JVM version, etc... and we'll try to loop in some of the OpenHFT folks and see if there's anything they can point us toward.
Update: It seems that the issue with the test is actually in the cleanup which may fail on some OSs if the file descriptors aren't correctly released. That is a benign error and one we'll try and get a good fix for. In the meantime, I'd say it's safe to add @Ignore
to the test and not worry that the PersistentQueue
stuff isn't working since it's just test cleanup that's failing and not the functionality of the Java Chronicle itself.