Please review at your convenience. Issue: https://bugs.openjdk.java.net/browse/JDK-8078586 Patch: http://cr.openjdk.java.net/~bpb/8078586/webrev.00/
Summary: Add setting of Random seed via system property with fallback to a random seed from a separate Random instance and printing the seed actually used in the test. The question is whether this change should be made under JDK-8078586 which is better resolved as “Cannot Reproduce” given its 0.07% failure rate, or whether a separate issue should be filed for these test changes. Thanks, Brian