I think that the changes to make 2nd-ary cache close invocations wait for close to complete is resulting in some intermittent suspect log statements:
[fatal 2020/05/01 18:25:41.450 GMT <org.apache.geode.distributed.internal.InternalLocator$$Lambda$306/0x0000000840ac5840@b563755> tid=90] Could not stop Locator on 0.0.0.0/0.0.0.0:40561:40561 in 120 seconds And it is true that InternalLocator close might have to wait a little longer for Cache close to complete. I'm assuming it takes longer only when something like a GC occurs. If anyone has any more data on this, please share! Thanks, Kirk
