Github user nickwallen commented on the issue:
https://github.com/apache/metron/pull/1015
Daft. Still a problem. In master, run repetitively the test fails easily
3/10 times for me. With the latest in this PR, it is failing 1/40 times or so,
so I'm still not there.
So even explicitly performing cache maintenance does not mean that the
cache size will always be exactly what we expect.
I am wondering if we even need such low level validation as what we have in
testWithFullCache() now.---
