chore: analyze Pod ReadinessProbe faliure

Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>
This commit is contained in:
Peter Pan
2023-04-19 23:43:40 -04:00
parent 0472c363a4
commit 3c7e0bba1d
2 changed files with 75 additions and 16 deletions

View File

@@ -44,7 +44,7 @@ func (PodAnalyzer) Analyze(a common.Analyzer) ([]common.Result, error) {
}
}
// Check through container status to check for crashes
// Check through container status to check for crashes or unready
for _, containerStatus := range pod.Status.ContainerStatuses {
if containerStatus.State.Waiting != nil {
if containerStatus.State.Waiting.Reason == "CrashLoopBackOff" || containerStatus.State.Waiting.Reason == "ImagePullBackOff" {
@@ -70,6 +70,23 @@ func (PodAnalyzer) Analyze(a common.Analyzer) ([]common.Result, error) {
})
}
}
} else {
// when pod is Running but its ReadinessProbe fails
if containerStatus.Ready == false && pod.Status.Phase == "Running" {
// parse the event log and append details
evt, err := FetchLatestEvent(a.Context, a.Client, pod.Namespace, pod.Name)
if err != nil || evt == nil {
continue
}
if evt.Reason == "Unhealthy" && evt.Message != "" {
failures = append(failures, common.Failure{
Text: evt.Message,
Sensitive: []common.Sensitive{},
})
}
}
}
}
if len(failures) > 0 {

View File

@@ -14,23 +14,65 @@ import (
func TestPodAnalyzer(t *testing.T) {
clientset := fake.NewSimpleClientset(&v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "example",
Namespace: "default",
Annotations: map[string]string{},
},
Status: v1.PodStatus{
Phase: v1.PodPending,
Conditions: []v1.PodCondition{
{
Type: v1.PodScheduled,
Reason: "Unschedulable",
Message: "0/1 nodes are available: 1 node(s) had taint {node-role.kubernetes.io/master: }, that the pod didn't tolerate.",
clientset := fake.NewSimpleClientset(
&v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "example",
Namespace: "default",
Annotations: map[string]string{},
},
Status: v1.PodStatus{
Phase: v1.PodPending,
Conditions: []v1.PodCondition{
{
Type: v1.PodScheduled,
Reason: "Unschedulable",
Message: "0/1 nodes are available: 1 node(s) had taint {node-role.kubernetes.io/master: }, that the pod didn't tolerate.",
},
},
},
},
})
&v1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: "example2",
Namespace: "default",
},
Status: v1.PodStatus{
Phase: v1.PodRunning,
ContainerStatuses: []v1.ContainerStatus{
{
Name: "example2",
Ready: false,
},
},
Conditions: []v1.PodCondition{
{
Type: v1.ContainersReady,
Reason: "ContainersNotReady",
Message: "containers with unready status: [example2]",
},
},
},
},
// simulate event: 30s Warning Unhealthy pod/my-nginx-7fb4dbcf47-4ch4w Readiness probe failed: bash: xxxx: command not found
&v1.Event{
ObjectMeta: metav1.ObjectMeta{
Name: "foo",
Namespace: "default",
},
InvolvedObject: v1.ObjectReference{
Kind: "Pod",
Name: "example2",
Namespace: "default",
UID: "differentUid",
APIVersion: "v1",
},
Reason: "Unhealthy",
Message: "readiness probe failed: the detail reason here ...",
Source: v1.EventSource{Component: "eventTest"},
Count: 1,
Type: v1.EventTypeWarning,
})
config := common.Analyzer{
Client: &kubernetes.Client{
@@ -45,7 +87,7 @@ func TestPodAnalyzer(t *testing.T) {
if err != nil {
t.Error(err)
}
assert.Equal(t, len(analysisResults), 1)
assert.Equal(t, len(analysisResults), 2)
}
func TestPodAnalyzerNamespaceFiltering(t *testing.T) {