diff --git a/pkg/apis/core/helper/qos/qos.go b/pkg/apis/core/helper/qos/qos.go
index d93553c6d20..6323122f98c 100644
--- a/pkg/apis/core/helper/qos/qos.go
+++ b/pkg/apis/core/helper/qos/qos.go
@@ -30,21 +30,21 @@ func isSupportedQoSComputeResource(name core.ResourceName) bool {
 	return supportedQoSComputeResources.Has(string(name))
 }
 
-// PodQOSClass returns the QoS class of a pod persisted in the PodStatus.
-// If QOSClass is empty, it returns value of GetPodQOS() which computes pod's QoS class.
-func PodQOSClass(pod *core.Pod) core.PodQOSClass {
+// GetPodQOS returns the QoS class of a pod persisted in the PodStatus.QOSClass field.
+// If PodStatus.QOSClass is empty, it returns value of ComputePodQOS() which evaluates pod's QoS class.
+func GetPodQOS(pod *core.Pod) core.PodQOSClass {
 	if pod.Status.QOSClass != "" {
 		return pod.Status.QOSClass
 	}
-	return GetPodQOS(pod)
+	return ComputePodQOS(pod)
 }
 
-// GetPodQOS returns the QoS class of a pod.
+// ComputePodQOS evaluates the list of containers to determine a pod's QoS class. This function is expensive.
 // A pod is besteffort if none of its containers have specified any requests or limits.
 // A pod is guaranteed only when requests and limits are specified for all the containers and they are equal.
 // A pod is burstable if limits and requests do not match across all containers.
 // When this function is updated please also update staging/src/k8s.io/kubectl/pkg/util/qos/qos.go
-func GetPodQOS(pod *core.Pod) core.PodQOSClass {
+func ComputePodQOS(pod *core.Pod) core.PodQOSClass {
 	requests := core.ResourceList{}
 	limits := core.ResourceList{}
 	zeroQuantity := resource.MustParse("0")
diff --git a/pkg/apis/core/v1/helper/qos/qos.go b/pkg/apis/core/v1/helper/qos/qos.go
index 7dde74e8644..e0db483dcca 100644
--- a/pkg/apis/core/v1/helper/qos/qos.go
+++ b/pkg/apis/core/v1/helper/qos/qos.go
@@ -32,20 +32,20 @@ func isSupportedQoSComputeResource(name v1.ResourceName) bool {
 	return supportedQoSComputeResources.Has(string(name))
 }
 
-// PodQOSClass returns the QoS class of a pod persisted in the PodStatus.
-// If QOSClass is empty, it returns value of GetPodQOS() which computes pod's QoS class.
-func PodQOSClass(pod *v1.Pod) v1.PodQOSClass {
+// GetPodQOS returns the QoS class of a pod persisted in the PodStatus.QOSClass field.
+// If PodStatus.QOSClass is empty, it returns value of ComputePodQOS() which evaluates pod's QoS class.
+func GetPodQOS(pod *v1.Pod) v1.PodQOSClass {
 	if pod.Status.QOSClass != "" {
 		return pod.Status.QOSClass
 	}
-	return GetPodQOS(pod)
+	return ComputePodQOS(pod)
 }
 
-// GetPodQOS returns the QoS class of a pod.
+// ComputePodQOS evaluates the list of containers to determine a pod's QoS class. This function is expensive.
 // A pod is besteffort if none of its containers have specified any requests or limits.
 // A pod is guaranteed only when requests and limits are specified for all the containers and they are equal.
 // A pod is burstable if limits and requests do not match across all containers.
-func GetPodQOS(pod *v1.Pod) v1.PodQOSClass {
+func ComputePodQOS(pod *v1.Pod) v1.PodQOSClass {
 	requests := v1.ResourceList{}
 	limits := v1.ResourceList{}
 	zeroQuantity := resource.MustParse("0")
diff --git a/pkg/apis/core/v1/helper/qos/qos_test.go b/pkg/apis/core/v1/helper/qos/qos_test.go
index 6dd45f61438..d16c17a14e7 100644
--- a/pkg/apis/core/v1/helper/qos/qos_test.go
+++ b/pkg/apis/core/v1/helper/qos/qos_test.go
@@ -27,7 +27,7 @@ import (
 	corev1 "k8s.io/kubernetes/pkg/apis/core/v1"
 )
 
-func TestGetPodQOS(t *testing.T) {
+func TestComputePodQOS(t *testing.T) {
 	testCases := []struct {
 		pod      *v1.Pod
 		expected v1.PodQOSClass
@@ -128,15 +128,15 @@ func TestGetPodQOS(t *testing.T) {
 		},
 	}
 	for id, testCase := range testCases {
-		if actual := GetPodQOS(testCase.pod); testCase.expected != actual {
+		if actual := ComputePodQOS(testCase.pod); testCase.expected != actual {
 			t.Errorf("[%d]: invalid qos pod %s, expected: %s, actual: %s", id, testCase.pod.Name, testCase.expected, actual)
 		}
 
-		// Convert v1.Pod to core.Pod, and then check against `core.helper.GetPodQOS`.
+		// Convert v1.Pod to core.Pod, and then check against `core.helper.ComputePodQOS`.
 		pod := core.Pod{}
 		corev1.Convert_v1_Pod_To_core_Pod(testCase.pod, &pod, nil)
 
-		if actual := qos.GetPodQOS(&pod); core.PodQOSClass(testCase.expected) != actual {
+		if actual := qos.ComputePodQOS(&pod); core.PodQOSClass(testCase.expected) != actual {
 			t.Errorf("[%d]: conversion invalid qos pod %s, expected: %s, actual: %s", id, testCase.pod.Name, testCase.expected, actual)
 		}
 	}
diff --git a/pkg/apis/core/validation/validation.go b/pkg/apis/core/validation/validation.go
index fdf28401352..b39e59bc9d9 100644
--- a/pkg/apis/core/validation/validation.go
+++ b/pkg/apis/core/validation/validation.go
@@ -4785,7 +4785,7 @@ func ValidatePodUpdate(newPod, oldPod *core.Pod, opts PodValidationOptions) fiel
 		return allErrs
 	}
 
-	if qos.PodQOSClass(oldPod) != qos.GetPodQOS(newPod) {
+	if qos.GetPodQOS(oldPod) != qos.ComputePodQOS(newPod) {
 		allErrs = append(allErrs, field.Invalid(fldPath, newPod.Status.QOSClass, "Pod QoS is immutable"))
 	}
 
diff --git a/pkg/kubelet/cm/cpumanager/policy_static.go b/pkg/kubelet/cm/cpumanager/policy_static.go
index b5ba3c96b48..0f72e64dbc8 100644
--- a/pkg/kubelet/cm/cpumanager/policy_static.go
+++ b/pkg/kubelet/cm/cpumanager/policy_static.go
@@ -419,7 +419,7 @@ func (p *staticPolicy) allocateCPUs(s state.State, numCPUs int, numaAffinity bit
 }
 
 func (p *staticPolicy) guaranteedCPUs(pod *v1.Pod, container *v1.Container) int {
-	if v1qos.PodQOSClass(pod) != v1.PodQOSGuaranteed {
+	if v1qos.GetPodQOS(pod) != v1.PodQOSGuaranteed {
 		return 0
 	}
 	cpuQuantity := container.Resources.Requests[v1.ResourceCPU]
diff --git a/pkg/kubelet/cm/helpers_linux.go b/pkg/kubelet/cm/helpers_linux.go
index 99fb2f055a2..8a144e7a73c 100644
--- a/pkg/kubelet/cm/helpers_linux.go
+++ b/pkg/kubelet/cm/helpers_linux.go
@@ -165,7 +165,7 @@ func ResourceConfigForPod(pod *v1.Pod, enforceCPULimits bool, cpuPeriod uint64,
 	}
 
 	// determine the qos class
-	qosClass := v1qos.PodQOSClass(pod)
+	qosClass := v1qos.GetPodQOS(pod)
 
 	// build the result
 	result := &ResourceConfig{}
diff --git a/pkg/kubelet/cm/memorymanager/policy_static.go b/pkg/kubelet/cm/memorymanager/policy_static.go
index daa2597ea2b..54391235e60 100644
--- a/pkg/kubelet/cm/memorymanager/policy_static.go
+++ b/pkg/kubelet/cm/memorymanager/policy_static.go
@@ -93,7 +93,7 @@ func (p *staticPolicy) Start(s state.State) error {
 // Allocate call is idempotent
 func (p *staticPolicy) Allocate(s state.State, pod *v1.Pod, container *v1.Container) error {
 	// allocate the memory only for guaranteed pods
-	if v1qos.PodQOSClass(pod) != v1.PodQOSGuaranteed {
+	if v1qos.GetPodQOS(pod) != v1.PodQOSGuaranteed {
 		return nil
 	}
 
@@ -362,7 +362,7 @@ func getPodRequestedResources(pod *v1.Pod) (map[v1.ResourceName]uint64, error) {
 }
 
 func (p *staticPolicy) GetPodTopologyHints(s state.State, pod *v1.Pod) map[string][]topologymanager.TopologyHint {
-	if v1qos.PodQOSClass(pod) != v1.PodQOSGuaranteed {
+	if v1qos.GetPodQOS(pod) != v1.PodQOSGuaranteed {
 		return nil
 	}
 
@@ -390,7 +390,7 @@ func (p *staticPolicy) GetPodTopologyHints(s state.State, pod *v1.Pod) map[strin
 // and is consulted to achieve NUMA aware resource alignment among this
 // and other resource controllers.
 func (p *staticPolicy) GetTopologyHints(s state.State, pod *v1.Pod, container *v1.Container) map[string][]topologymanager.TopologyHint {
-	if v1qos.PodQOSClass(pod) != v1.PodQOSGuaranteed {
+	if v1qos.GetPodQOS(pod) != v1.PodQOSGuaranteed {
 		return nil
 	}
 
diff --git a/pkg/kubelet/cm/pod_container_manager_linux.go b/pkg/kubelet/cm/pod_container_manager_linux.go
index 268ba92bb4b..dae7f8bd3d4 100644
--- a/pkg/kubelet/cm/pod_container_manager_linux.go
+++ b/pkg/kubelet/cm/pod_container_manager_linux.go
@@ -99,7 +99,7 @@ func (m *podContainerManagerImpl) EnsureExists(pod *v1.Pod) error {
 
 // GetPodContainerName returns the CgroupName identifier, and its literal cgroupfs form on the host.
 func (m *podContainerManagerImpl) GetPodContainerName(pod *v1.Pod) (CgroupName, string) {
-	podQOS := v1qos.PodQOSClass(pod)
+	podQOS := v1qos.GetPodQOS(pod)
 	// Get the parent QOS container name
 	var parentContainer CgroupName
 	switch podQOS {
diff --git a/pkg/kubelet/cm/qos_container_manager_linux.go b/pkg/kubelet/cm/qos_container_manager_linux.go
index 91deddded30..abf4487ee5d 100644
--- a/pkg/kubelet/cm/qos_container_manager_linux.go
+++ b/pkg/kubelet/cm/qos_container_manager_linux.go
@@ -173,7 +173,7 @@ func (m *qosContainerManagerImpl) setCPUCgroupConfig(configs map[v1.PodQOSClass]
 	reuseReqs := make(v1.ResourceList, 4)
 	for i := range pods {
 		pod := pods[i]
-		qosClass := v1qos.PodQOSClass(pod)
+		qosClass := v1qos.GetPodQOS(pod)
 		if qosClass != v1.PodQOSBurstable {
 			// we only care about the burstable qos tier
 			continue
@@ -207,7 +207,7 @@ func (m *qosContainerManagerImpl) getQoSMemoryRequests() map[v1.PodQOSClass]int6
 	reuseReqs := make(v1.ResourceList, 4)
 	for _, pod := range pods {
 		podMemoryRequest := int64(0)
-		qosClass := v1qos.PodQOSClass(pod)
+		qosClass := v1qos.GetPodQOS(pod)
 		if qosClass == v1.PodQOSBestEffort {
 			// limits are not set for Best Effort pods
 			continue
diff --git a/pkg/kubelet/eviction/eviction_manager.go b/pkg/kubelet/eviction/eviction_manager.go
index c29b47ffcd3..e47b37a0d05 100644
--- a/pkg/kubelet/eviction/eviction_manager.go
+++ b/pkg/kubelet/eviction/eviction_manager.go
@@ -151,7 +151,7 @@ func (m *managerImpl) Admit(attrs *lifecycle.PodAdmitAttributes) lifecycle.PodAd
 	// Conditions other than memory pressure reject all pods
 	nodeOnlyHasMemoryPressureCondition := hasNodeCondition(m.nodeConditions, v1.NodeMemoryPressure) && len(m.nodeConditions) == 1
 	if nodeOnlyHasMemoryPressureCondition {
-		notBestEffort := v1.PodQOSBestEffort != v1qos.PodQOSClass(attrs.Pod)
+		notBestEffort := v1.PodQOSBestEffort != v1qos.GetPodQOS(attrs.Pod)
 		if notBestEffort {
 			return lifecycle.PodAdmitResult{Admit: true}
 		}
diff --git a/pkg/kubelet/kubelet_pods.go b/pkg/kubelet/kubelet_pods.go
index ea5f9ee28b0..03d8dc987b2 100644
--- a/pkg/kubelet/kubelet_pods.go
+++ b/pkg/kubelet/kubelet_pods.go
@@ -1839,7 +1839,7 @@ func (kl *Kubelet) convertStatusToAPIStatus(pod *v1.Pod, podStatus *kubecontaine
 	}
 
 	// set status for Pods created on versions of kube older than 1.6
-	apiPodStatus.QOSClass = v1qos.PodQOSClass(pod)
+	apiPodStatus.QOSClass = v1qos.GetPodQOS(pod)
 
 	apiPodStatus.ContainerStatuses = kl.convertToAPIContainerStatuses(
 		pod, podStatus,
diff --git a/pkg/kubelet/preemption/preemption.go b/pkg/kubelet/preemption/preemption.go
index 92f379537fb..e4d0cbd931b 100644
--- a/pkg/kubelet/preemption/preemption.go
+++ b/pkg/kubelet/preemption/preemption.go
@@ -247,7 +247,7 @@ func (a admissionRequirementList) toString() string {
 func sortPodsByQOS(preemptor *v1.Pod, pods []*v1.Pod) (bestEffort, burstable, guaranteed []*v1.Pod) {
 	for _, pod := range pods {
 		if kubetypes.Preemptable(preemptor, pod) {
-			switch v1qos.PodQOSClass(pod) {
+			switch v1qos.GetPodQOS(pod) {
 			case v1.PodQOSBestEffort:
 				bestEffort = append(bestEffort, pod)
 			case v1.PodQOSBurstable:
diff --git a/pkg/kubelet/qos/policy.go b/pkg/kubelet/qos/policy.go
index fea10465e9c..7117be21255 100644
--- a/pkg/kubelet/qos/policy.go
+++ b/pkg/kubelet/qos/policy.go
@@ -46,7 +46,7 @@ func GetContainerOOMScoreAdjust(pod *v1.Pod, container *v1.Container, memoryCapa
 		return guaranteedOOMScoreAdj
 	}
 
-	switch v1qos.PodQOSClass(pod) {
+	switch v1qos.GetPodQOS(pod) {
 	case v1.PodQOSGuaranteed:
 		// Guaranteed containers should be the last to get killed.
 		return guaranteedOOMScoreAdj
diff --git a/pkg/quota/v1/evaluator/core/pods.go b/pkg/quota/v1/evaluator/core/pods.go
index 90bbd96880e..a201c58383e 100644
--- a/pkg/quota/v1/evaluator/core/pods.go
+++ b/pkg/quota/v1/evaluator/core/pods.go
@@ -375,7 +375,7 @@ func PodUsageFunc(obj runtime.Object, clock clock.Clock) (corev1.ResourceList, e
 }
 
 func isBestEffort(pod *corev1.Pod) bool {
-	return qos.PodQOSClass(pod) == corev1.PodQOSBestEffort
+	return qos.GetPodQOS(pod) == corev1.PodQOSBestEffort
 }
 
 func isTerminating(pod *corev1.Pod) bool {
diff --git a/plugin/pkg/admission/podtolerationrestriction/admission.go b/plugin/pkg/admission/podtolerationrestriction/admission.go
index 7dcf7c8690b..4e20ee96558 100644
--- a/plugin/pkg/admission/podtolerationrestriction/admission.go
+++ b/plugin/pkg/admission/podtolerationrestriction/admission.go
@@ -99,7 +99,7 @@ func (p *Plugin) Admit(ctx context.Context, a admission.Attributes, o admission.
 		extraTolerations = ts
 	}
 
-	if qoshelper.PodQOSClass(pod) != api.PodQOSBestEffort {
+	if qoshelper.GetPodQOS(pod) != api.PodQOSBestEffort {
 		extraTolerations = append(extraTolerations, api.Toleration{
 			Key:      corev1.TaintNodeMemoryPressure,
 			Operator: api.TolerationOpExists,
diff --git a/staging/src/k8s.io/kubectl/pkg/describe/describe.go b/staging/src/k8s.io/kubectl/pkg/describe/describe.go
index faf3696344c..7b88f0f9184 100644
--- a/staging/src/k8s.io/kubectl/pkg/describe/describe.go
+++ b/staging/src/k8s.io/kubectl/pkg/describe/describe.go
@@ -871,7 +871,7 @@ func describePod(pod *corev1.Pod, events *corev1.EventList) (string, error) {
 			}
 		}
 		describeVolumes(pod.Spec.Volumes, w, "")
-		w.Write(LEVEL_0, "QoS Class:\t%s\n", qos.PodQOSClass(pod))
+		w.Write(LEVEL_0, "QoS Class:\t%s\n", qos.GetPodQOS(pod))
 		printLabelsMultiline(w, "Node-Selectors", pod.Spec.NodeSelector)
 		printPodTolerationsMultiline(w, "Tolerations", pod.Spec.Tolerations)
 		describeTopologySpreadConstraints(pod.Spec.TopologySpreadConstraints, w, "")
diff --git a/staging/src/k8s.io/kubectl/pkg/util/qos/qos.go b/staging/src/k8s.io/kubectl/pkg/util/qos/qos.go
index 87183a095a6..16a8846a799 100644
--- a/staging/src/k8s.io/kubectl/pkg/util/qos/qos.go
+++ b/staging/src/k8s.io/kubectl/pkg/util/qos/qos.go
@@ -28,20 +28,20 @@ func isSupportedQoSComputeResource(name core.ResourceName) bool {
 	return supportedQoSComputeResources.Has(string(name))
 }
 
-// PodQOSClass returns the QoS class of a pod persisted in the PodStatus.
-// If QOSClass is empty, it returns value of GetPodQOS() which computes pod's QoS class.
-func PodQOSClass(pod *core.Pod) core.PodQOSClass {
+// GetPodQOS returns the QoS class of a pod persisted in the PodStatus.QOSClass field.
+// If PodStatus.QOSClass is empty, it returns value of ComputePodQOS() which evaluates pod's QoS class.
+func GetPodQOS(pod *core.Pod) core.PodQOSClass {
 	if pod.Status.QOSClass != "" {
 		return pod.Status.QOSClass
 	}
-	return GetPodQOS(pod)
+	return ComputePodQOS(pod)
 }
 
-// GetPodQOS returns the QoS class of a pod.
+// ComputePodQOS evaluates the list of containers to determine a pod's QoS class. This function is expensive.
 // A pod is besteffort if none of its containers have specified any requests or limits.
 // A pod is guaranteed only when requests and limits are specified for all the containers and they are equal.
 // A pod is burstable if limits and requests do not match across all containers.
-func GetPodQOS(pod *core.Pod) core.PodQOSClass {
+func ComputePodQOS(pod *core.Pod) core.PodQOSClass {
 	requests := core.ResourceList{}
 	limits := core.ResourceList{}
 	zeroQuantity := resource.MustParse("0")
diff --git a/test/e2e/scheduling/priorities.go b/test/e2e/scheduling/priorities.go
index ff0cfd78592..63197bc67fa 100644
--- a/test/e2e/scheduling/priorities.go
+++ b/test/e2e/scheduling/priorities.go
@@ -501,7 +501,7 @@ func computeCPUMemFraction(node v1.Node, resource *v1.ResourceRequirements, pods
 	for _, pod := range pods {
 		framework.Logf("Pod for on the node: %v, Cpu: %v, Mem: %v", pod.Name, getNonZeroRequests(pod).MilliCPU, getNonZeroRequests(pod).Memory)
 		// Ignore best effort pods while computing fractions as they won't be taken in account by scheduler.
-		if v1qos.PodQOSClass(pod) == v1.PodQOSBestEffort {
+		if v1qos.GetPodQOS(pod) == v1.PodQOSBestEffort {
 			continue
 		}
 		totalRequestedCPUResource += getNonZeroRequests(pod).MilliCPU