Closed crazyyanchao closed 4 years ago
@crazyyanchao Did you perform an upgrade on one of the nodes and then copy that upgrade node to be the database of the other nodes in the cluster?
@bradnussbaum Yes,I do this.But 3.5.16 is ok! Why is it happen?
@crazyyanchao Does it happen from a clean startup of 3.6.0.M1? After you upgraded in SINGLE mode from 3.5.16 to 3.6.0.M1 did you allow the node to come fully online? After the node came online and you shutdown cleanly, did you run neo4j-admin unbind
to clear the cluster state and then neo4j-admin load
to load the upgraded graph into each node?
@bradnussbaum:) Thanks for your reply. Yes,it happen from a clean startup of 3.6.0.M1.But I haven't try to run unbind and load.I will try this.But is that a necessary step for 3.6.*? Now I still shift to 3.5.16.It is ok.I will continue to test...