mirror of
https://github.com/kubernetes/kubernetes.git
synced 2026-06-23 16:32:02 -04:00
10efa46fbb
The Memory Manager Metrics BeforeEach asserts that zero pods are running on the node after a kubelet config update. This hard assertion flakes when a preceding serial test's namespace deletion hasn't completed yet — framework namespace cleanup is async and the kubelet restart in updateKubeletConfig can delay in-flight pod termination. CI logs show leftover pods from MemoryQoS tests (memqos-burstable, memqos-no-limit, etc.), Probe Stress tests (50-container pods), and Summary API PSI tests (memory-pressure-pod), all still Running when the assertion fires 4-7ms after the previous test finishes. Replace the immediate Expect(count).To(BeZero()) with an Eventually poll (2 minute timeout, 5 second interval) that gives pods time to drain after the kubelet restart. The existing printAllPodsOnNode diagnostic output is preserved inside the poll for debugging. Signed-off-by: Davanum Srinivas <davanum@gmail.com>
255 lines
10 KiB
Go
255 lines
10 KiB
Go
//go:build linux
|
|
|
|
/*
|
|
Copyright 2023 The Kubernetes Authors.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
*/
|
|
|
|
package e2enode
|
|
|
|
import (
|
|
"context"
|
|
"fmt"
|
|
"time"
|
|
|
|
"github.com/onsi/ginkgo/v2"
|
|
"github.com/onsi/gomega"
|
|
"github.com/onsi/gomega/gstruct"
|
|
v1 "k8s.io/api/core/v1"
|
|
"k8s.io/apimachinery/pkg/api/resource"
|
|
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
|
|
"k8s.io/apimachinery/pkg/fields"
|
|
clientset "k8s.io/client-go/kubernetes"
|
|
kubeletconfig "k8s.io/kubernetes/pkg/kubelet/apis/config"
|
|
"k8s.io/kubernetes/test/e2e/feature"
|
|
"k8s.io/kubernetes/test/e2e/framework"
|
|
e2emetrics "k8s.io/kubernetes/test/e2e/framework/metrics"
|
|
e2epod "k8s.io/kubernetes/test/e2e/framework/pod"
|
|
admissionapi "k8s.io/pod-security-admission/api"
|
|
)
|
|
|
|
var _ = SIGDescribe("Memory Manager Metrics", framework.WithSerial(), feature.MemoryManager, func() {
|
|
f := framework.NewDefaultFramework("memorymanager-metrics")
|
|
f.NamespacePodSecurityLevel = admissionapi.LevelPrivileged
|
|
|
|
ginkgo.Context("when querying /metrics", func() {
|
|
var oldCfg *kubeletconfig.KubeletConfiguration
|
|
var testPod *v1.Pod
|
|
|
|
ginkgo.BeforeEach(func(ctx context.Context) {
|
|
var err error
|
|
if oldCfg == nil {
|
|
oldCfg, err = getCurrentKubeletConfig(ctx)
|
|
framework.ExpectNoError(err)
|
|
}
|
|
|
|
newCfg := oldCfg.DeepCopy()
|
|
updateKubeletConfigWithMemoryManagerParams(newCfg,
|
|
&memoryManagerKubeletParams{
|
|
policy: staticPolicy,
|
|
systemReservedMemory: []kubeletconfig.MemoryReservation{
|
|
{
|
|
NumaNode: 0,
|
|
Limits: v1.ResourceList{
|
|
resourceMemory: resource.MustParse("1100Mi"),
|
|
},
|
|
},
|
|
},
|
|
systemReserved: map[string]string{resourceMemory: "500Mi"},
|
|
kubeReserved: map[string]string{resourceMemory: "500Mi"},
|
|
evictionHard: map[string]string{evictionHardMemory: "100Mi"},
|
|
},
|
|
)
|
|
updateKubeletConfig(ctx, f, newCfg, true)
|
|
ginkgo.DeferCleanup(func(ctx context.Context) {
|
|
if testPod != nil {
|
|
deletePodSyncByName(ctx, f, testPod.Name)
|
|
}
|
|
updateKubeletConfig(ctx, f, oldCfg, true)
|
|
})
|
|
|
|
// Wait for pods from previous serial tests to finish terminating.
|
|
// Framework namespace deletion is async, and the kubelet restart
|
|
// above can delay in-flight pod cleanup. Poll instead of asserting
|
|
// immediately so we don't flake on test ordering.
|
|
gomega.Eventually(ctx, func(ctx context.Context) error {
|
|
count, err := printAllPodsOnNode(ctx, f.ClientSet, framework.TestContext.NodeName)
|
|
if err != nil {
|
|
return err
|
|
}
|
|
if count != 0 {
|
|
return fmt.Errorf("unexpected pods on %q: %d still present", framework.TestContext.NodeName, count)
|
|
}
|
|
return nil
|
|
}).WithTimeout(2*time.Minute).WithPolling(5*time.Second).Should(
|
|
gomega.Succeed(), "unexpected pods on %q, please check output above", framework.TestContext.NodeName)
|
|
})
|
|
|
|
ginkgo.It("should report zero pinning counters after a fresh restart", func(ctx context.Context) {
|
|
// we updated the kubelet config in BeforeEach, so we can assume we start fresh.
|
|
// being [Serial], we can also assume no one else but us is running pods.
|
|
ginkgo.By("Checking the memorymanager metrics right after the kubelet restart, with no pods running")
|
|
|
|
matchResourceMetrics := gstruct.MatchKeys(gstruct.IgnoreExtras, gstruct.Keys{
|
|
"kubelet_memory_manager_pinning_requests_total": gstruct.MatchAllElements(nodeID, gstruct.Elements{
|
|
"": timelessSample(0), // intentionally use stricter value
|
|
}),
|
|
"kubelet_memory_manager_pinning_errors_total": gstruct.MatchAllElements(nodeID, gstruct.Elements{
|
|
"": timelessSample(0), // intentionally use stricter value
|
|
}),
|
|
})
|
|
|
|
ginkgo.By("Giving the Kubelet time to start up and produce metrics")
|
|
gomega.Eventually(getKubeletMetrics, 1*time.Minute, 15*time.Second).WithContext(ctx).Should(matchResourceMetrics)
|
|
ginkgo.By("Ensuring the metrics match the expectations a few more times")
|
|
gomega.Consistently(getKubeletMetrics, 1*time.Minute, 15*time.Second).WithContext(ctx).Should(matchResourceMetrics)
|
|
})
|
|
|
|
ginkgo.It("should report pinning failures when the memorymanager allocation is known to fail", func(ctx context.Context) {
|
|
ginkgo.By("Creating the test pod which will be rejected for memory request which is too big")
|
|
testPod = e2epod.NewPodClient(f).Create(ctx, makeMemoryManagerPod("memmngrpod", nil,
|
|
[]memoryManagerCtnAttributes{
|
|
{
|
|
ctnName: "memmngrcnt",
|
|
cpus: "100m",
|
|
memory: "1000Gi"},
|
|
}),
|
|
)
|
|
|
|
// we updated the kubelet config in BeforeEach, so we can assume we start fresh.
|
|
// being [Serial], we can also assume noone else but us is running pods.
|
|
ginkgo.By("Checking the memorymanager metrics right after the kubelet restart, with pod failed to admit")
|
|
|
|
matchResourceMetrics := gstruct.MatchKeys(gstruct.IgnoreExtras, gstruct.Keys{
|
|
"kubelet_memory_manager_pinning_requests_total": gstruct.MatchAllElements(nodeID, gstruct.Elements{
|
|
"": timelessSample(1),
|
|
}),
|
|
"kubelet_memory_manager_pinning_errors_total": gstruct.MatchAllElements(nodeID, gstruct.Elements{
|
|
"": timelessSample(1),
|
|
}),
|
|
})
|
|
|
|
ginkgo.By("Giving the Kubelet time to start up and produce metrics")
|
|
gomega.Eventually(getKubeletMetrics, 1*time.Minute, 15*time.Second).WithContext(ctx).Should(matchResourceMetrics)
|
|
ginkgo.By("Ensuring the metrics match the expectations a few more times")
|
|
gomega.Consistently(getKubeletMetrics, 1*time.Minute, 15*time.Second).WithContext(ctx).Should(matchResourceMetrics)
|
|
|
|
values, err := getKubeletMetrics(ctx)
|
|
framework.ExpectNoError(err, "error getting the kubelet metrics for sanity check")
|
|
err = validateMetrics(
|
|
values,
|
|
"kubelet_memory_manager_pinning_requests_total",
|
|
"kubelet_memory_manager_pinning_errors_total",
|
|
func(totVal, errVal float64) error {
|
|
if int64(totVal) != int64(errVal) {
|
|
return fmt.Errorf("expected total requests equal to total errors")
|
|
}
|
|
return nil
|
|
},
|
|
)
|
|
framework.ExpectNoError(err, "error validating the kubelet metrics between each other")
|
|
})
|
|
|
|
ginkgo.It("should not report any pinning failures when the memorymanager allocation is expected to succeed", func(ctx context.Context) {
|
|
ginkgo.By("Creating the test pod")
|
|
testPod = e2epod.NewPodClient(f).Create(ctx, makeMemoryManagerPod("memmngrpod", nil,
|
|
[]memoryManagerCtnAttributes{
|
|
{
|
|
ctnName: "memmngrcnt",
|
|
cpus: "100m",
|
|
memory: "64Mi",
|
|
},
|
|
}),
|
|
)
|
|
|
|
_, err := printAllPodsOnNode(ctx, f.ClientSet, framework.TestContext.NodeName)
|
|
framework.ExpectNoError(err)
|
|
|
|
// we updated the kubelet config in BeforeEach, so we can assume we start fresh.
|
|
// being [Serial], we can also assume noone else but us is running pods.
|
|
ginkgo.By("Checking the memorymanager metrics right after the kubelet restart, with pod should be admitted")
|
|
matchResourceMetrics := gstruct.MatchKeys(gstruct.IgnoreExtras, gstruct.Keys{
|
|
"kubelet_memory_manager_pinning_requests_total": gstruct.MatchAllElements(nodeID, gstruct.Elements{
|
|
"": timelessSample(1),
|
|
}),
|
|
"kubelet_memory_manager_pinning_errors_total": gstruct.MatchAllElements(nodeID, gstruct.Elements{
|
|
"": timelessSample(0),
|
|
}),
|
|
})
|
|
|
|
ginkgo.By("Giving the Kubelet time to start up and produce metrics")
|
|
gomega.Eventually(getKubeletMetrics, 1*time.Minute, 15*time.Second).WithContext(ctx).Should(matchResourceMetrics)
|
|
ginkgo.By("Ensuring the metrics match the expectations a few more times")
|
|
gomega.Consistently(getKubeletMetrics, 1*time.Minute, 15*time.Second).WithContext(ctx).Should(matchResourceMetrics)
|
|
|
|
values, err := getKubeletMetrics(ctx)
|
|
framework.ExpectNoError(err, "error getting the kubelet metrics for sanity check")
|
|
err = validateMetrics(
|
|
values,
|
|
"kubelet_memory_manager_pinning_requests_total",
|
|
"kubelet_memory_manager_pinning_errors_total",
|
|
func(totVal, errVal float64) error {
|
|
if int64(totVal-errVal) < 1 {
|
|
return fmt.Errorf("expected total requests equal to total errors + 1")
|
|
}
|
|
return nil
|
|
},
|
|
)
|
|
framework.ExpectNoError(err, "error validating the kubelet metrics between each other")
|
|
})
|
|
})
|
|
})
|
|
|
|
func validateMetrics(values e2emetrics.KubeletMetrics, totalKey, errorKey string, checkFn func(totVal, errVal float64) error) error {
|
|
totalSamples := values[totalKey]
|
|
errorSamples := values[errorKey]
|
|
if len(totalSamples) != len(errorSamples) {
|
|
return fmt.Errorf("inconsistent samples, total=%d error=%d", len(totalSamples), len(errorSamples))
|
|
}
|
|
for idx := range totalSamples {
|
|
if err := checkFn(float64(totalSamples[idx].Value), float64(errorSamples[idx].Value)); err != nil {
|
|
return err
|
|
}
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// printAllPodsOnNode outputs status of all kubelet pods into log.
|
|
// Note considering the e2e_node environment we will always have exactly 1 node, but still.
|
|
func printAllPodsOnNode(ctx context.Context, c clientset.Interface, nodeName string) (int, error) {
|
|
nodeSelector := fields.Set{
|
|
"spec.nodeName": nodeName,
|
|
}.AsSelector().String()
|
|
|
|
podList, err := c.CoreV1().Pods(metav1.NamespaceAll).List(ctx, metav1.ListOptions{
|
|
FieldSelector: nodeSelector,
|
|
})
|
|
if err != nil {
|
|
framework.Logf("Unable to retrieve pods for node %v: %v", nodeName, err)
|
|
return 0, err
|
|
}
|
|
count := 0
|
|
framework.Logf("begin listing pods: %d found", len(podList.Items))
|
|
for _, p := range podList.Items {
|
|
framework.Logf("%s/%s node %s (expected: %s) status %v QoS %s message %s reason %s (%d container statuses recorded)",
|
|
p.Namespace, p.Name, p.Spec.NodeName, nodeName, p.Status.Phase, p.Status.QOSClass, p.Status.Message, p.Status.Reason, len(p.Status.ContainerStatuses))
|
|
for _, c := range p.Status.ContainerStatuses {
|
|
framework.Logf("\tContainer %v ready: %v, restart count %v",
|
|
c.Name, c.Ready, c.RestartCount)
|
|
}
|
|
count++
|
|
}
|
|
framework.Logf("end listing pods: %d found", len(podList.Items))
|
|
return count, nil
|
|
}
|