diff --git a/cluster/calcium/realloc.go b/cluster/calcium/realloc.go
index 6b57e6745..80ddb1b39 100644
--- a/cluster/calcium/realloc.go
+++ b/cluster/calcium/realloc.go
@@ -29,6 +29,7 @@ func (c *Calcium) ReallocResource(ctx context.Context, opts *types.ReallocOption
 				StorageLimit:    workload.StorageLimit + opts.ResourceOpts.StorageLimit,
 				VolumeRequest:   types.MergeVolumeBindings(workload.VolumeRequest, opts.ResourceOpts.VolumeRequest),
 				VolumeLimit:     types.MergeVolumeBindings(workload.VolumeLimit, opts.ResourceOpts.VolumeLimit),
+				VolumeExist:     workload.VolumePlanRequest,
 			},
 		)
 		if err != nil {
@@ -57,8 +58,7 @@ func (c *Calcium) doReallocOnNode(ctx context.Context, nodename string, workload
 				resourceMeta := &types.ResourceMeta{}
 				for _, plan := range plans {
 					if resourceMeta, err = plan.Dispense(resourcetypes.DispenseOptions{
-						Node:             node,
-						ExistingInstance: workload,
+						Node: node,
 					}, resourceMeta); err != nil {
 						return err
 					}
diff --git a/resources/types/types.go b/resources/types/types.go
index 9c82babc9..1291b6029 100644
--- a/resources/types/types.go
+++ b/resources/types/types.go
@@ -63,8 +63,7 @@ type SchedulerV2 func(context.Context, []ScheduleInfo) (ResourcePlans, int, erro
 // DispenseOptions .
 type DispenseOptions struct {
 	*types.Node
-	Index            int
-	ExistingInstance *types.Workload
+	Index int
 }
 
 // ResourcePlans .
diff --git a/resources/volume/volume.go b/resources/volume/volume.go
index 32b555d45..4f26ce7ae 100644
--- a/resources/volume/volume.go
+++ b/resources/volume/volume.go
@@ -17,6 +17,7 @@ type volumeRequest struct {
 	limit    [maxVolumes]types.VolumeBinding
 	requests int
 	limits   int
+	Existing types.VolumePlan
 }
 
 // MakeRequest .
@@ -30,6 +31,7 @@ func MakeRequest(opts types.ResourceOptions) (resourcetypes.ResourceRequest, err
 	}
 	v.requests = len(opts.VolumeRequest)
 	v.limits = len(opts.VolumeLimit)
+	v.Existing = opts.VolumeExist
 
 	sort.Slice(opts.VolumeLimit, func(i, j int) bool {
 		return opts.VolumeLimit[i].ToString(false) < opts.VolumeLimit[j].ToString(false)
@@ -87,7 +89,12 @@ func (v volumeRequest) MakeScheduler() resourcetypes.SchedulerV2 {
 			limit = append(limit, &v.limit[i])
 		}
 
-		scheduleInfos, volumePlans, total, err := schedulerV1.SelectVolumeNodes(ctx, scheduleInfos, request)
+		var volumePlans map[string][]types.VolumePlan
+		if v.Existing != nil {
+			scheduleInfos[0], volumePlans, total, err = schedulerV1.ReselectVolumeNodes(ctx, scheduleInfos[0], v.Existing, request)
+		} else {
+			scheduleInfos, volumePlans, total, err = schedulerV1.SelectVolumeNodes(ctx, scheduleInfos, request)
+		}
 		return ResourcePlans{
 			capacity: resourcetypes.GetCapacity(scheduleInfos),
 			request:  request,
@@ -172,22 +179,6 @@ func (rp ResourcePlans) Dispense(opts resourcetypes.DispenseOptions, r *types.Re
 	}
 	r.VolumePlanRequest = rp.plan[opts.Node.Name][opts.Index]
 
-	// if there are existing ones, ensure new volumes are compatible
-	if opts.ExistingInstance != nil {
-		found := false
-		for _, plan := range rp.plan[opts.Node.Name] {
-			if plan.Compatible(opts.ExistingInstance.VolumePlanRequest) {
-				r.VolumePlanRequest = plan
-				found = true
-				break
-			}
-		}
-
-		if !found {
-			return nil, errors.Wrap(types.ErrInsufficientVolume, "incompatible volume plans")
-		}
-	}
-
 	// fix plans while limit > request
 	r.VolumePlanLimit = types.VolumePlan{}
 	for i := range rp.request {
@@ -204,7 +195,8 @@ func (rp ResourcePlans) Dispense(opts resourcetypes.DispenseOptions, r *types.Re
 	}
 
 	// judge if volume changed
-	r.VolumeChanged = opts.ExistingInstance != nil && !r.VolumeLimit.IsEqual(opts.ExistingInstance.VolumeLimit)
+	// TODO@zc
+	r.VolumeChanged = false
 	return r, nil
 }
 
diff --git a/scheduler/complex/potassium.go b/scheduler/complex/potassium.go
index b9fdc0db7..5299b846b 100644
--- a/scheduler/complex/potassium.go
+++ b/scheduler/complex/potassium.go
@@ -364,6 +364,10 @@ func (m *Potassium) ReselectVolumeNodes(ctx context.Context, scheduleInfo resour
 	affinityPlan.Merge(unlimPlan)
 
 	// schedule new volume requests
+	if len(needReschedule) == 0 {
+		scheduleInfo.Capacity = 1
+		return scheduleInfo, map[string][]types.VolumePlan{scheduleInfo.Name: []types.VolumePlan{affinityPlan}}, 1, nil
+	}
 	scheduleInfos, volumePlans, total, err := m.SelectVolumeNodes(ctx, []resourcetypes.ScheduleInfo{scheduleInfo}, needReschedule)
 	if err != nil {
 		return scheduleInfo, nil, 0, err
diff --git a/types/resource.go b/types/resource.go
index 8635b40c2..2354e68e6 100644
--- a/types/resource.go
+++ b/types/resource.go
@@ -19,6 +19,7 @@ type ResourceOptions struct {
 
 	VolumeRequest VolumeBindings
 	VolumeLimit   VolumeBindings
+	VolumeExist   VolumePlan
 
 	StorageRequest int64
 	StorageLimit   int64
@@ -238,3 +239,14 @@ func (p VolumePlan) Merge(p2 VolumePlan) {
 		p[vb] = vm
 	}
 }
+
+// FindAffinityPlan .
+func (p VolumePlan) FindAffinityPlan(req VolumeBinding) (_ VolumeBinding, _ VolumeMap, found bool) {
+	for vb, vm := range p {
+		if vb.Source == req.Source && vb.Destination == req.Destination && vb.Flags == req.Flags {
+			return vb, vm, true
+		}
+	}
+	found = false
+	return
+}