At first if you are having only one backup you surely data lose when kill 3 nodes (you can to do that when only kill one by one and waiting rebalancing complete after each).
Could you please attache a full log file at leas one node where the remap failed messages are present? -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/what-does-remaps-mean-in-DataStreamerImpl-java-tp12033p12222.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
