forked from kubernetes/kubernetes
-
Notifications
You must be signed in to change notification settings - Fork 1
/
container_probe.go
448 lines (409 loc) · 14.1 KB
/
container_probe.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
/*
Copyright 2015 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 common
import (
"fmt"
"time"
"k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/util/intstr"
"k8s.io/apimachinery/pkg/util/uuid"
podutil "k8s.io/kubernetes/pkg/api/v1/pod"
"k8s.io/kubernetes/test/e2e/framework"
testutils "k8s.io/kubernetes/test/utils"
. "github.com/onsi/ginkgo"
. "github.com/onsi/gomega"
imageutils "k8s.io/kubernetes/test/utils/image"
)
const (
probTestContainerName = "test-webserver"
probTestInitialDelaySeconds = 15
defaultObservationTimeout = time.Minute * 2
)
var _ = framework.KubeDescribe("Probing container", func() {
f := framework.NewDefaultFramework("container-probe")
var podClient *framework.PodClient
probe := webserverProbeBuilder{}
BeforeEach(func() {
podClient = f.PodClient()
})
/*
Testname: pods-readiness-probe-initial-delay
Description: Make sure that pod with readiness probe should not be
ready before initial delay and never restart.
*/
framework.ConformanceIt("with readiness probe should not be ready before initial delay and never restart ", func() {
p := podClient.Create(makePodSpec(probe.withInitialDelay().build(), nil))
f.WaitForPodReady(p.Name)
p, err := podClient.Get(p.Name, metav1.GetOptions{})
framework.ExpectNoError(err)
isReady, err := testutils.PodRunningReady(p)
framework.ExpectNoError(err)
Expect(isReady).To(BeTrue(), "pod should be ready")
// We assume the pod became ready when the container became ready. This
// is true for a single container pod.
readyTime, err := getTransitionTimeForReadyCondition(p)
framework.ExpectNoError(err)
startedTime, err := getContainerStartedTime(p, probTestContainerName)
framework.ExpectNoError(err)
framework.Logf("Container started at %v, pod became ready at %v", startedTime, readyTime)
initialDelay := probTestInitialDelaySeconds * time.Second
if readyTime.Sub(startedTime) < initialDelay {
framework.Failf("Pod became ready before it's %v initial delay", initialDelay)
}
restartCount := getRestartCount(p)
Expect(restartCount == 0).To(BeTrue(), "pod should have a restart count of 0 but got %v", restartCount)
})
/*
Testname: pods-readiness-probe-failure
Description: Make sure that pod with readiness probe that fails should
never be ready and never restart.
*/
framework.ConformanceIt("with readiness probe that fails should never be ready and never restart ", func() {
p := podClient.Create(makePodSpec(probe.withFailing().build(), nil))
Consistently(func() (bool, error) {
p, err := podClient.Get(p.Name, metav1.GetOptions{})
if err != nil {
return false, err
}
return podutil.IsPodReady(p), nil
}, 1*time.Minute, 1*time.Second).ShouldNot(BeTrue(), "pod should not be ready")
p, err := podClient.Get(p.Name, metav1.GetOptions{})
framework.ExpectNoError(err)
isReady, err := testutils.PodRunningReady(p)
Expect(isReady).NotTo(BeTrue(), "pod should be not ready")
restartCount := getRestartCount(p)
Expect(restartCount == 0).To(BeTrue(), "pod should have a restart count of 0 but got %v", restartCount)
})
/*
Testname: pods-cat-liveness-probe-restarted
Description: Make sure the pod is restarted with a cat /tmp/health
liveness probe.
*/
framework.ConformanceIt("should be restarted with a exec \"cat /tmp/health\" liveness probe", func() {
runLivenessTest(f, &v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "liveness-exec",
Labels: map[string]string{"test": "liveness"},
},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: "liveness",
Image: busyboxImage,
Command: []string{"/bin/sh", "-c", "echo ok >/tmp/health; sleep 10; rm -rf /tmp/health; sleep 600"},
LivenessProbe: &v1.Probe{
Handler: v1.Handler{
Exec: &v1.ExecAction{
Command: []string{"cat", "/tmp/health"},
},
},
InitialDelaySeconds: 15,
FailureThreshold: 1,
},
},
},
},
}, 1, defaultObservationTimeout)
})
/*
Testname: pods-cat-liveness-probe-not-restarted
Description: Make sure the pod is not restarted with a cat /tmp/health
liveness probe.
*/
framework.ConformanceIt("should *not* be restarted with a exec \"cat /tmp/health\" liveness probe", func() {
runLivenessTest(f, &v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "liveness-exec",
Labels: map[string]string{"test": "liveness"},
},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: "liveness",
Image: busyboxImage,
Command: []string{"/bin/sh", "-c", "echo ok >/tmp/health; sleep 600"},
LivenessProbe: &v1.Probe{
Handler: v1.Handler{
Exec: &v1.ExecAction{
Command: []string{"cat", "/tmp/health"},
},
},
InitialDelaySeconds: 15,
FailureThreshold: 1,
},
},
},
},
}, 0, defaultObservationTimeout)
})
/*
Testname: pods-http-liveness-probe-restarted
Description: Make sure when http liveness probe fails, the pod should
be restarted.
*/
framework.ConformanceIt("should be restarted with a /healthz http liveness probe ", func() {
runLivenessTest(f, &v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "liveness-http",
Labels: map[string]string{"test": "liveness"},
},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: "liveness",
Image: imageutils.GetE2EImage(imageutils.Liveness),
Command: []string{"/server"},
LivenessProbe: &v1.Probe{
Handler: v1.Handler{
HTTPGet: &v1.HTTPGetAction{
Path: "/healthz",
Port: intstr.FromInt(8080),
},
},
InitialDelaySeconds: 15,
FailureThreshold: 1,
},
},
},
},
}, 1, defaultObservationTimeout)
})
// Slow by design (5 min)
/*
Testname: pods-restart-count
Description: Make sure when a pod gets restarted, its start count
should increase.
*/
framework.ConformanceIt("should have monotonically increasing restart count [Slow]", func() {
runLivenessTest(f, &v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "liveness-http",
Labels: map[string]string{"test": "liveness"},
},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: "liveness",
Image: imageutils.GetE2EImage(imageutils.Liveness),
Command: []string{"/server"},
LivenessProbe: &v1.Probe{
Handler: v1.Handler{
HTTPGet: &v1.HTTPGetAction{
Path: "/healthz",
Port: intstr.FromInt(8080),
},
},
InitialDelaySeconds: 5,
FailureThreshold: 1,
},
},
},
},
}, 5, time.Minute*5)
})
/*
Testname: pods-http-liveness-probe-not-restarted
Description: Make sure when http liveness probe succeeds, the pod
should not be restarted.
*/
framework.ConformanceIt("should *not* be restarted with a /healthz http liveness probe ", func() {
runLivenessTest(f, &v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "liveness-http",
Labels: map[string]string{"test": "liveness"},
},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: "liveness",
Image: imageutils.GetE2EImage(imageutils.NginxSlim),
Ports: []v1.ContainerPort{{ContainerPort: 80}},
LivenessProbe: &v1.Probe{
Handler: v1.Handler{
HTTPGet: &v1.HTTPGetAction{
Path: "/",
Port: intstr.FromInt(80),
},
},
InitialDelaySeconds: 15,
TimeoutSeconds: 5,
FailureThreshold: 5, // to accommodate nodes which are slow in bringing up containers.
},
},
},
},
}, 0, defaultObservationTimeout)
})
/*
Testname: pods-docker-liveness-probe-timeout
Description: Make sure that the pod is restarted with a docker exec
liveness probe with timeout.
*/
framework.ConformanceIt("should be restarted with a docker exec liveness probe with timeout ", func() {
// TODO: enable this test once the default exec handler supports timeout.
framework.Skipf("The default exec handler, dockertools.NativeExecHandler, does not support timeouts due to a limitation in the Docker Remote API")
runLivenessTest(f, &v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "liveness-exec",
Labels: map[string]string{"test": "liveness"},
},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: "liveness",
Image: busyboxImage,
Command: []string{"/bin/sh", "-c", "sleep 600"},
LivenessProbe: &v1.Probe{
Handler: v1.Handler{
Exec: &v1.ExecAction{
Command: []string{"/bin/sh", "-c", "sleep 10"},
},
},
InitialDelaySeconds: 15,
TimeoutSeconds: 1,
FailureThreshold: 1,
},
},
},
},
}, 1, defaultObservationTimeout)
})
})
func getContainerStartedTime(p *v1.Pod, containerName string) (time.Time, error) {
for _, status := range p.Status.ContainerStatuses {
if status.Name != containerName {
continue
}
if status.State.Running == nil {
return time.Time{}, fmt.Errorf("Container is not running")
}
return status.State.Running.StartedAt.Time, nil
}
return time.Time{}, fmt.Errorf("cannot find container named %q", containerName)
}
func getTransitionTimeForReadyCondition(p *v1.Pod) (time.Time, error) {
for _, cond := range p.Status.Conditions {
if cond.Type == v1.PodReady {
return cond.LastTransitionTime.Time, nil
}
}
return time.Time{}, fmt.Errorf("No ready condition can be found for pod")
}
func getRestartCount(p *v1.Pod) int {
count := 0
for _, containerStatus := range p.Status.ContainerStatuses {
count += int(containerStatus.RestartCount)
}
return count
}
func makePodSpec(readinessProbe, livenessProbe *v1.Probe) *v1.Pod {
pod := &v1.Pod{
ObjectMeta: metav1.ObjectMeta{Name: "test-webserver-" + string(uuid.NewUUID())},
Spec: v1.PodSpec{
Containers: []v1.Container{
{
Name: probTestContainerName,
Image: imageutils.GetE2EImage(imageutils.TestWebserver),
LivenessProbe: livenessProbe,
ReadinessProbe: readinessProbe,
},
},
},
}
return pod
}
type webserverProbeBuilder struct {
failing bool
initialDelay bool
}
func (b webserverProbeBuilder) withFailing() webserverProbeBuilder {
b.failing = true
return b
}
func (b webserverProbeBuilder) withInitialDelay() webserverProbeBuilder {
b.initialDelay = true
return b
}
func (b webserverProbeBuilder) build() *v1.Probe {
probe := &v1.Probe{
Handler: v1.Handler{
HTTPGet: &v1.HTTPGetAction{
Port: intstr.FromInt(80),
Path: "/",
},
},
}
if b.initialDelay {
probe.InitialDelaySeconds = probTestInitialDelaySeconds
}
if b.failing {
probe.HTTPGet.Port = intstr.FromInt(81)
}
return probe
}
func runLivenessTest(f *framework.Framework, pod *v1.Pod, expectNumRestarts int, timeout time.Duration) {
podClient := f.PodClient()
ns := f.Namespace.Name
Expect(pod.Spec.Containers).NotTo(BeEmpty())
containerName := pod.Spec.Containers[0].Name
// At the end of the test, clean up by removing the pod.
defer func() {
By("deleting the pod")
podClient.Delete(pod.Name, metav1.NewDeleteOptions(0))
}()
By(fmt.Sprintf("Creating pod %s in namespace %s", pod.Name, ns))
podClient.Create(pod)
// Wait until the pod is not pending. (Here we need to check for something other than
// 'Pending' other than checking for 'Running', since when failures occur, we go to
// 'Terminated' which can cause indefinite blocking.)
framework.ExpectNoError(framework.WaitForPodNotPending(f.ClientSet, ns, pod.Name),
fmt.Sprintf("starting pod %s in namespace %s", pod.Name, ns))
framework.Logf("Started pod %s in namespace %s", pod.Name, ns)
// Check the pod's current state and verify that restartCount is present.
By("checking the pod's current state and verifying that restartCount is present")
pod, err := podClient.Get(pod.Name, metav1.GetOptions{})
framework.ExpectNoError(err, fmt.Sprintf("getting pod %s in namespace %s", pod.Name, ns))
initialRestartCount := podutil.GetExistingContainerStatus(pod.Status.ContainerStatuses, containerName).RestartCount
framework.Logf("Initial restart count of pod %s is %d", pod.Name, initialRestartCount)
// Wait for the restart state to be as desired.
deadline := time.Now().Add(timeout)
lastRestartCount := initialRestartCount
observedRestarts := int32(0)
for start := time.Now(); time.Now().Before(deadline); time.Sleep(2 * time.Second) {
pod, err = podClient.Get(pod.Name, metav1.GetOptions{})
framework.ExpectNoError(err, fmt.Sprintf("getting pod %s", pod.Name))
restartCount := podutil.GetExistingContainerStatus(pod.Status.ContainerStatuses, containerName).RestartCount
if restartCount != lastRestartCount {
framework.Logf("Restart count of pod %s/%s is now %d (%v elapsed)",
ns, pod.Name, restartCount, time.Since(start))
if restartCount < lastRestartCount {
framework.Failf("Restart count should increment monotonically: restart cont of pod %s/%s changed from %d to %d",
ns, pod.Name, lastRestartCount, restartCount)
}
}
observedRestarts = restartCount - initialRestartCount
if expectNumRestarts > 0 && int(observedRestarts) >= expectNumRestarts {
// Stop if we have observed more than expectNumRestarts restarts.
break
}
lastRestartCount = restartCount
}
// If we expected 0 restarts, fail if observed any restart.
// If we expected n restarts (n > 0), fail if we observed < n restarts.
if (expectNumRestarts == 0 && observedRestarts > 0) || (expectNumRestarts > 0 &&
int(observedRestarts) < expectNumRestarts) {
framework.Failf("pod %s/%s - expected number of restarts: %d, found restarts: %d",
ns, pod.Name, expectNumRestarts, observedRestarts)
}
}