Cluster doesn't recognize forcibly (windows equivalent to kill -9) killed node on windows. I tried it with OPENIDM 6.0.0-RC10 and external dj 6.0.0-RC3 as repo on windows server 2012. Also I tried it with automated test with mssql db. So it looks as environment problem because on linux it works OK.
How to reproduce
1. install cluster with 2 nodes with external dj as repo by doc
2. run dj and nodes with default conf folder
3. kill one node (I killed through windows Task Manager)
4. wait 1 minute and then check cluster state by rest or by ui
and we get this:
How you can see both nodes states are running but node1 in my cause was killed ..