Fix for "Process repair-queue"

Signed-off-by: morimoto-cybozu <[email protected]>

op/repair_drain_start.go

@@ -38,7 +38,7 @@ func (o *repairDrainStartOp) NextCommand() cke.Commander {
 
 	attempts := 1
 	if o.config.EvictRetries != nil {
-		attempts = *o.config.EvictRetries
+		attempts = *o.config.EvictRetries + 1
 	}
 	interval := 0 * time.Second
 	if o.config.EvictInterval != nil {

phase.go

@@ -18,7 +18,7 @@ const (
 	PhaseEtcdMaintain    = OperationPhase("etcd-maintain")
 	PhaseK8sMaintain     = OperationPhase("k8s-maintain")
 	PhaseStopCP          = OperationPhase("stop-control-plane")
-	PhaseRepairNodes     = OperationPhase("repair-nodes")
+	PhaseRepairMachines  = OperationPhase("repair-machines")
 	PhaseUncordonNodes   = OperationPhase("uncordon-nodes")
 	PhaseRebootNodes     = OperationPhase("reboot-nodes")
 	PhaseCompleted       = OperationPhase("completed")
@@ -37,7 +37,7 @@ var AllOperationPhases = []OperationPhase{
 	PhaseEtcdMaintain,
 	PhaseK8sMaintain,
 	PhaseStopCP,
-	PhaseRepairNodes,
+	PhaseRepairMachines,
 	PhaseUncordonNodes,
 	PhaseRebootNodes,
 	PhaseCompleted,

server/strategy.go

@@ -97,9 +97,9 @@ func DecideOps(c *cke.Cluster, cs *cke.ClusterStatus, constraints *cke.Constrain
 	if ops, phaseRepair := repairOps(c, cs, constraints, rebootArgs, nf); phaseRepair {
 		if !nf.EtcdIsGood() {
 			log.Warn("cannot repair machines because etcd cluster is not responding and in-sync", nil)
-			return nil, cke.PhaseRepairNodes
+			return nil, cke.PhaseRepairMachines
 		}
-		return ops, cke.PhaseRepairNodes
+		return ops, cke.PhaseRepairMachines
 	}
 
 	// 11. Reboot nodes if reboot request has been arrived to the reboot queue, and the number of unreachable nodes is less than a threshold.