-
Notifications
You must be signed in to change notification settings - Fork 74
/
scheduler.go
196 lines (170 loc) · 6.86 KB
/
scheduler.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
package scheduler
import (
"fmt"
"os"
"time"
v1alpha1 "github.com/hwameistor/hwameistor/pkg/apis/hwameistor/v1alpha1"
lvmscheduler "github.com/hwameistor/hwameistor/pkg/local-storage/member/controller/scheduler"
log "github.com/sirupsen/logrus"
corev1 "k8s.io/api/core/v1"
corev1lister "k8s.io/client-go/listers/core/v1"
storagev1lister "k8s.io/client-go/listers/storage/v1"
"k8s.io/client-go/rest"
"k8s.io/klog"
framework "k8s.io/kubernetes/pkg/scheduler/framework/v1alpha1"
"sigs.k8s.io/controller-runtime/pkg/manager"
"sigs.k8s.io/controller-runtime/pkg/manager/signals"
)
// VolumeScheduler is to scheduler hwameistor volume
type Scheduler struct {
lvmScheduler VolumeScheduler
diskScheduler VolumeScheduler
pvLister corev1lister.PersistentVolumeLister
pvcLister corev1lister.PersistentVolumeClaimLister
scLister storagev1lister.StorageClassLister
}
// NewDataCache creates a cache instance
func NewScheduler(f framework.FrameworkHandle) *Scheduler {
cfg, err := rest.InClusterConfig()
if err != nil {
log.WithError(err).Fatal("Failed to construct the cluster config")
}
options := manager.Options{
MetricsBindAddress: "0", // disable metrics
}
mgr, err := manager.New(cfg, options)
if err != nil {
klog.V(1).Info(err)
os.Exit(1)
}
// Setup Scheme for all resources of Local Storage
if err := v1alpha1.AddToScheme(mgr.GetScheme()); err != nil {
log.WithError(err).Fatal("Failed to setup scheme for local-storage resources")
}
apiClient := mgr.GetClient()
hwameiStorCache := mgr.GetCache()
sche := Scheduler{
pvLister: f.SharedInformerFactory().Core().V1().PersistentVolumes().Lister(),
pvcLister: f.SharedInformerFactory().Core().V1().PersistentVolumeClaims().Lister(),
scLister: f.SharedInformerFactory().Storage().V1().StorageClasses().Lister(),
}
ctx := signals.SetupSignalHandler()
go func() {
mgr.Start(ctx)
}()
replicaScheduler := lvmscheduler.New(apiClient, hwameiStorCache, 1000)
// wait for cache synced
for {
if hwameiStorCache.WaitForCacheSync(ctx) {
break
}
time.Sleep(time.Second * 1)
}
replicaScheduler.Init()
sche.lvmScheduler = NewLVMVolumeScheduler(f, replicaScheduler, hwameiStorCache, apiClient)
sche.diskScheduler = NewDiskVolumeScheduler(f)
return &sche
}
func (s *Scheduler) Reserve(pod *corev1.Pod, node string) error {
_, _, _, diskNewPVCs, err := s.getHwameiStorPVCs(pod)
if err != nil {
return err
}
return s.diskScheduler.Reserve(diskNewPVCs, node)
}
func (s *Scheduler) Unreserve(pod *corev1.Pod, node string) error {
_, _, _, diskNewPVCs, err := s.getHwameiStorPVCs(pod)
if err != nil {
return err
}
return s.diskScheduler.Unreserve(diskNewPVCs, node)
}
func (s *Scheduler) Filter(pod *corev1.Pod, node *corev1.Node) (bool, error) {
lvmProvisionedPVCs, lvmNewPVCs, diskProvisionedPVCs, diskNewPVCs, err := s.getHwameiStorPVCs(pod)
if err != nil {
return false, err
}
// figure out the existing local volume associated to the PVC, and send it to the scheduler's filter
existingLocalVolumes := []string{}
for _, pvc := range lvmProvisionedPVCs {
pv, err := s.pvLister.Get(pvc.Spec.VolumeName)
if err != nil {
log.WithFields(log.Fields{"pvc": pvc.Name, "namespace": pvc.Namespace, "pv": pvc.Spec.VolumeName}).WithError(err).Error("Failed to get a Provisioned PVC's PV")
return false, err
}
if pv.Spec.CSI == nil || len(pv.Spec.CSI.VolumeHandle) == 0 {
log.WithFields(log.Fields{"pvc": pvc.Name, "namespace": pvc.Namespace, "pv": pvc.Spec.VolumeName}).Error("Wrong PV status of a Provisioned PVC")
return false, fmt.Errorf("wrong pv")
}
existingLocalVolumes = append(existingLocalVolumes, pv.Spec.CSI.VolumeHandle)
}
canSchedule, err := s.lvmScheduler.Filter(existingLocalVolumes, lvmNewPVCs, node)
if err != nil {
return false, err
}
if !canSchedule {
return false, fmt.Errorf("can't schedule the LVM volume to node %s", node.Name)
}
// figure out the existing local volume associated to the PVC, and send it to the scheduler's filter
existingLocalVolumes = []string{}
for _, pvc := range diskProvisionedPVCs {
pv, err := s.pvLister.Get(pvc.Spec.VolumeName)
if err != nil {
log.WithFields(log.Fields{"pvc": pvc.Name, "namespace": pvc.Namespace, "pv": pvc.Spec.VolumeName}).WithError(err).Error("Failed to get a Provisioned PVC's PV")
return false, err
}
if pv.Spec.CSI == nil || len(pv.Spec.CSI.VolumeHandle) == 0 {
log.WithFields(log.Fields{"pvc": pvc.Name, "namespace": pvc.Namespace, "pv": pvc.Spec.VolumeName}).Error("Wrong PV status of a Provisioned PVC")
return false, fmt.Errorf("wrong pv")
}
existingLocalVolumes = append(existingLocalVolumes, pv.Spec.CSI.VolumeHandle)
}
return s.diskScheduler.Filter(existingLocalVolumes, diskNewPVCs, node)
}
// return: lvmProvisionedClaims, lvmNewClaims, diskProvisionedClaims, diskNewClaims, error
func (s *Scheduler) getHwameiStorPVCs(pod *corev1.Pod) ([]*corev1.PersistentVolumeClaim, []*corev1.PersistentVolumeClaim, []*corev1.PersistentVolumeClaim, []*corev1.PersistentVolumeClaim, error) {
lvmProvisionedClaims := []*corev1.PersistentVolumeClaim{}
lvmNewClaims := []*corev1.PersistentVolumeClaim{}
diskProvisionedClaims := []*corev1.PersistentVolumeClaim{}
diskNewClaims := []*corev1.PersistentVolumeClaim{}
lvmCSIDriverName := s.lvmScheduler.CSIDriverName()
diskCSIDriverName := s.diskScheduler.CSIDriverName()
for _, vol := range pod.Spec.Volumes {
if vol.PersistentVolumeClaim == nil {
continue
}
pvc, err := s.pvcLister.PersistentVolumeClaims(pod.Namespace).Get(vol.PersistentVolumeClaim.ClaimName)
if err != nil {
// if pvc can't be found in the cluster, the pod should not be able to be scheduled
return lvmProvisionedClaims, lvmNewClaims, diskProvisionedClaims, diskNewClaims, err
}
if pvc.Spec.StorageClassName == nil {
// should not be the CSI pvc, ignore
continue
}
sc, err := s.scLister.Get(*pvc.Spec.StorageClassName)
if err != nil {
// can't found storageclass in the cluster, the pod should not be able to be scheduled
return lvmProvisionedClaims, lvmNewClaims, diskProvisionedClaims, diskNewClaims, err
}
if sc.Provisioner == lvmCSIDriverName {
if pvc.Status.Phase == corev1.ClaimBound {
lvmProvisionedClaims = append(lvmProvisionedClaims, pvc)
} else if pvc.Status.Phase == corev1.ClaimPending {
lvmNewClaims = append(lvmNewClaims, pvc)
} else {
return lvmProvisionedClaims, lvmNewClaims, diskProvisionedClaims, diskNewClaims, fmt.Errorf("unhealthy HwameiStor LVM pvc")
}
}
if sc.Provisioner == diskCSIDriverName {
if pvc.Status.Phase == corev1.ClaimBound {
diskProvisionedClaims = append(diskProvisionedClaims, pvc)
} else if pvc.Status.Phase == corev1.ClaimPending {
diskNewClaims = append(diskNewClaims, pvc)
} else {
return lvmProvisionedClaims, lvmNewClaims, diskProvisionedClaims, diskNewClaims, fmt.Errorf("unhealthy HwameiStor Disk pvc")
}
}
}
return lvmProvisionedClaims, lvmNewClaims, diskProvisionedClaims, diskNewClaims, nil
}