Remove Exiting/Down node from other DC, #24171

* When leaving/downing the last node in a DC it would not
  be removed in another DC, since that was only done by the
  leader in the owning DC (and that is gone).
* It should be ok to eagerly remove such nodes also by
  leaders in other DCs.
* Note that gossip is already sent out so for the last node
  that will be spread to other DC, unless there is a network
  partition. For that we can't do anything. It will be replaced
  if joining again.

akka-cluster/src/multi-jvm/scala/akka/cluster/MultiDcLastNodeSpec.scala

@@ -0,0 +1,72 @@
+/**
+ * Copyright (C) 2018 Lightbend Inc. <https://www.lightbend.com>
+ */
+package akka.cluster
+
+import akka.remote.testkit.{ MultiNodeConfig, MultiNodeSpec }
+import com.typesafe.config.ConfigFactory
+
+import scala.concurrent.duration._
+
+object MultiDcLastNodeSpec extends MultiNodeConfig {
+  val first = role("first")
+  val second = role("second")
+  val third = role("third")
+
+  commonConfig(ConfigFactory.parseString(
+    s"""
+      #akka.loglevel = DEBUG
+    """).withFallback(MultiNodeClusterSpec.clusterConfig))
+
+  nodeConfig(first, second)(ConfigFactory.parseString(
+    """
+      akka.cluster.multi-data-center.self-data-center = "dc1"
+    """))
+
+  nodeConfig(third)(ConfigFactory.parseString(
+    """
+      akka.cluster.multi-data-center.self-data-center = "dc2"
+    """))
+
+}
+
+class MultiDcLastNodeMultiJvmNode1 extends MultiDcLastNodeSpec
+class MultiDcLastNodeMultiJvmNode2 extends MultiDcLastNodeSpec
+class MultiDcLastNodeMultiJvmNode3 extends MultiDcLastNodeSpec
+
+abstract class MultiDcLastNodeSpec extends MultiNodeSpec(MultiDcLastNodeSpec)
+  with MultiNodeClusterSpec {
+
+  import MultiDcLastNodeSpec._
+
+  "A multi-dc cluster with one remaining node in other DC" must {
+    "join" in {
+
+      runOn(first) {
+        cluster.join(first)
+      }
+      runOn(second, third) {
+        cluster.join(first)
+      }
+      enterBarrier("join-cluster")
+
+      within(20.seconds) {
+        awaitAssert(clusterView.members.filter(_.status == MemberStatus.Up) should have size 3)
+      }
+
+      enterBarrier("cluster started")
+    }
+
+    "be able to leave" in {
+      runOn(third) {
+        // this works in same way for down
+        cluster.leave(address(third))
+      }
+
+      runOn(first, second) {
+        awaitAssert(clusterView.members.map(_.address) should not contain address(third))
+      }
+      enterBarrier("cross-data-center-left")
+    }
+  }
+}