-
Notifications
You must be signed in to change notification settings - Fork 1.3k
/
node.go
318 lines (275 loc) · 9.47 KB
/
node.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
package watch
import (
"encoding/json"
"fmt"
"strings"
"time"
v1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/fields"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/types"
"k8s.io/apimachinery/pkg/util/wait"
"k8s.io/client-go/tools/cache"
"k8s.io/client-go/tools/record"
"k8s.io/client-go/util/workqueue"
virtv1 "kubevirt.io/kubevirt/pkg/api/v1"
"kubevirt.io/kubevirt/pkg/controller"
"kubevirt.io/kubevirt/pkg/kubecli"
"kubevirt.io/kubevirt/pkg/log"
)
const (
// NodeUnresponsiveReason is in various places as reason to indicate that
// an action was taken because virt-handler became unresponsive.
NodeUnresponsiveReason = "NodeUnresponsive"
)
// NodeController is the main NodeController struct.
type NodeController struct {
clientset kubecli.KubevirtClient
Queue workqueue.RateLimitingInterface
nodeInformer cache.SharedIndexInformer
vmiInformer cache.SharedIndexInformer
recorder record.EventRecorder
heartBeatTimeout time.Duration
recheckInterval time.Duration
}
// NewNodeController creates a new instance of the NodeController struct.
func NewNodeController(clientset kubecli.KubevirtClient, nodeInformer cache.SharedIndexInformer, vmiInformer cache.SharedIndexInformer, recorder record.EventRecorder) *NodeController {
c := &NodeController{
clientset: clientset,
Queue: workqueue.NewRateLimitingQueue(workqueue.DefaultControllerRateLimiter()),
nodeInformer: nodeInformer,
vmiInformer: vmiInformer,
recorder: recorder,
heartBeatTimeout: 5 * time.Minute,
recheckInterval: 1 * time.Minute,
}
c.nodeInformer.AddEventHandler(cache.ResourceEventHandlerFuncs{
AddFunc: c.addNode,
DeleteFunc: c.deleteNode,
UpdateFunc: c.updateNode,
})
c.vmiInformer.AddEventHandler(cache.ResourceEventHandlerFuncs{
AddFunc: c.addVirtualMachine,
DeleteFunc: func(_ interface{}) {}, // nothing to do
UpdateFunc: c.updateVirtualMachine,
})
return c
}
func (c *NodeController) addNode(obj interface{}) {
c.enqueueNode(obj)
}
func (c *NodeController) deleteNode(obj interface{}) {
c.enqueueNode(obj)
}
func (c *NodeController) updateNode(old, curr interface{}) {
c.enqueueNode(curr)
}
func (c *NodeController) enqueueNode(obj interface{}) {
logger := log.Log
node := obj.(*v1.Node)
key, err := controller.KeyFunc(node)
if err != nil {
logger.Object(node).Reason(err).Error("Failed to extract key from node.")
}
c.Queue.Add(key)
}
func (c *NodeController) addVirtualMachine(obj interface{}) {
vmi := obj.(*virtv1.VirtualMachineInstance)
if vmi.Status.NodeName != "" {
c.Queue.Add(vmi.Status.NodeName)
}
}
func (c *NodeController) updateVirtualMachine(old, curr interface{}) {
currVMI := curr.(*virtv1.VirtualMachineInstance)
if currVMI.Status.NodeName != "" {
c.Queue.Add(currVMI.Status.NodeName)
}
}
// Run runs the passed in NodeController.
func (c *NodeController) Run(threadiness int, stopCh chan struct{}) {
defer controller.HandlePanic()
defer c.Queue.ShutDown()
log.Log.Info("Starting node controller.")
// Wait for cache sync before we start the node controller
cache.WaitForCacheSync(stopCh, c.nodeInformer.HasSynced, c.vmiInformer.HasSynced)
// Start the actual work
for i := 0; i < threadiness; i++ {
go wait.Until(c.runWorker, time.Second, stopCh)
}
<-stopCh
log.Log.Info("Stopping node controller.")
}
func (c *NodeController) runWorker() {
for c.Execute() {
}
}
// Execute runs commands from the controller queue, if there is
// an error it requeues the command. Returns false if the queue
// is empty.
func (c *NodeController) Execute() bool {
key, quit := c.Queue.Get()
if quit {
return false
}
defer c.Queue.Done(key)
err := c.execute(key.(string))
if err != nil {
log.Log.Reason(err).Infof("reenqueuing node %v", key)
c.Queue.AddRateLimited(key)
} else {
log.Log.V(4).Infof("processed node %v", key)
c.Queue.Forget(key)
}
return true
}
func (c *NodeController) execute(key string) error {
obj, nodeExists, err := c.nodeInformer.GetStore().GetByKey(key)
logger := log.DefaultLogger()
if err != nil {
return err
}
nodeName := key
var node *v1.Node
if nodeExists {
node = obj.(*v1.Node)
logger = logger.Object(node)
} else {
logger = logger.Key(key, "Node")
}
if unresponsive, err := isNodeUnresponsive(node, c.heartBeatTimeout); err != nil {
logger.Reason(err).Error("Failed to dermine if node is responsive, will not reenqueue")
return fmt.Errorf("failed to determine if node %s is responsive: %v", nodeName, err)
} else if unresponsive {
if nodeExists && node.Labels[virtv1.NodeSchedulable] == "true" {
c.recorder.Event(node, v1.EventTypeNormal, NodeUnresponsiveReason, "virt-handler is not responsive, marking node as unresponsive")
data := []byte(fmt.Sprintf(`{"metadata": { "labels": {"%s": "false"}}}`, virtv1.NodeSchedulable))
_, err = c.clientset.CoreV1().Nodes().Patch(nodeName, types.StrategicMergePatchType, data)
if err != nil {
logger.Reason(err).Error("Failed to mark node as unschedulable")
return fmt.Errorf("failed to mark node %s as unschedulable: %v", nodeName, err)
}
}
vmis, err := c.virtualMachinesOnNode(nodeName)
if err != nil {
logger.Reason(err).Error("Failed fetch vmis for node")
return err
} else if len(vmis) == 0 {
if nodeExists {
c.Queue.AddAfter(key, c.recheckInterval)
}
return nil
}
pods, err := c.alivePodsOnNode(nodeName)
if err != nil {
logger.Reason(err).Error("Failed fetch pods for node")
return err
}
vmis = filterStuckVirtualMachinesWithoutPods(vmis, pods)
errs := []string{}
// Do sequential updates, we don't want to create update storms in situations where something might already be wrong
for _, vmi := range vmis {
c.recorder.Event(vmi, v1.EventTypeNormal, NodeUnresponsiveReason, fmt.Sprintf("virt-handler on node %s is not responsive, marking VMI as failed", vmi.Status.NodeName))
logger.V(2).Infof("Moving vmi %s in namespace %s on unresponsive node to failed state", vmi.Name, vmi.Namespace)
phasePatch := fmt.Sprintf(`{ "op": "replace", "path": "/status/phase", "value": "%s" }`, virtv1.Failed)
operation := "add"
if vmi.Status.Reason != "" {
operation = "replace"
}
reasonPatch := fmt.Sprintf(`{ "op": "%s", "path": "/status/reason", "value": "%s" }`, operation, NodeUnresponsiveReason)
_, err := c.clientset.VirtualMachineInstance(vmi.Namespace).Patch(vmi.Name, types.JSONPatchType, []byte(fmt.Sprintf("[%s, %s]", phasePatch, reasonPatch)))
if err != nil {
errs = append(errs, fmt.Sprintf("failed to move vmi %s in namespace %s to final state: %v", vmi.Name, vmi.Namespace, err))
logger.Reason(err).Errorf("Failed to move vmi %s in namespace %s to final state", vmi.Name, vmi.Namespace)
}
}
if len(errs) > 0 {
return fmt.Errorf("%v", strings.Join(errs, "; "))
}
}
if nodeExists {
c.Queue.AddAfter(key, c.recheckInterval)
}
return nil
}
func (c *NodeController) virtualMachinesOnNode(nodeName string) ([]*virtv1.VirtualMachineInstance, error) {
labelSelector, err := labels.Parse(fmt.Sprintf("%s in (%s)", virtv1.NodeNameLabel, nodeName))
if err != nil {
return nil, err
}
list, err := c.clientset.VirtualMachineInstance(v1.NamespaceAll).List(&metav1.ListOptions{
LabelSelector: labelSelector.String(),
})
if err != nil {
return nil, err
}
vmis := []*virtv1.VirtualMachineInstance{}
for i := range list.Items {
vmis = append(vmis, &list.Items[i])
}
return vmis, nil
}
func (c *NodeController) alivePodsOnNode(nodeName string) ([]*v1.Pod, error) {
handlerNodeSelector := fields.ParseSelectorOrDie("spec.nodeName=" + nodeName)
list, err := c.clientset.CoreV1().Pods(v1.NamespaceAll).List(metav1.ListOptions{
FieldSelector: handlerNodeSelector.String(),
})
if err != nil {
return nil, err
}
pods := []*v1.Pod{}
for i := range list.Items {
pod := &list.Items[i]
if controllerRef := controller.GetControllerOf(pod); !isControlledByVMI(controllerRef) {
continue
}
phase := pod.Status.Phase
if phase != v1.PodFailed && phase != v1.PodSucceeded {
pods = append(pods, pod)
}
}
return pods, nil
}
func filterStuckVirtualMachinesWithoutPods(vmis []*virtv1.VirtualMachineInstance, pods []*v1.Pod) []*virtv1.VirtualMachineInstance {
podsPerNamespace := map[string]map[string]*v1.Pod{}
for _, pod := range pods {
podsForVMI, ok := podsPerNamespace[pod.Namespace]
if !ok {
podsForVMI = map[string]*v1.Pod{}
}
if controllerRef := controller.GetControllerOf(pod); isControlledByVMI(controllerRef) {
podsForVMI[string(controllerRef.UID)] = pod
podsPerNamespace[pod.Namespace] = podsForVMI
}
}
filtered := []*virtv1.VirtualMachineInstance{}
for _, vmi := range vmis {
if vmi.IsScheduled() || vmi.IsRunning() {
if podsForVMI, exists := podsPerNamespace[vmi.Namespace]; exists {
if _, exists := podsForVMI[string(vmi.UID)]; exists {
continue
}
}
filtered = append(filtered, vmi)
}
}
return filtered
}
func isControlledByVMI(controllerRef *metav1.OwnerReference) bool {
return controllerRef != nil && controllerRef.Kind == virtv1.VirtualMachineInstanceGroupVersionKind.Kind
}
func isNodeUnresponsive(node *v1.Node, timeout time.Duration) (bool, error) {
if node == nil {
return true, nil
}
if lastHeartBeat, exists := node.Annotations[virtv1.VirtHandlerHeartbeat]; exists {
timestamp := metav1.Time{}
if err := json.Unmarshal([]byte(`"`+lastHeartBeat+`"`), ×tamp); err != nil {
return false, err
}
if timestamp.Time.Before(metav1.Now().Add(-timeout)) {
return true, nil
}
}
return false, nil
}