-
Notifications
You must be signed in to change notification settings - Fork 36
/
statefulset.go
1643 lines (1369 loc) · 46.1 KB
/
statefulset.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
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
package controllers
import (
"context"
"fmt"
"reflect"
"sort"
"strconv"
"strings"
"time"
appsv1 "k8s.io/api/apps/v1"
corev1 "k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/types"
"k8s.io/apimachinery/pkg/util/intstr"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/client-go/util/retry"
"sigs.k8s.io/controller-runtime/pkg/client"
"sigs.k8s.io/controller-runtime/pkg/controller/controllerutil"
asdbv1 "github.com/aerospike/aerospike-kubernetes-operator/api/v1"
"github.com/aerospike/aerospike-kubernetes-operator/pkg/utils"
lib "github.com/aerospike/aerospike-management-lib"
)
const (
// aerospike-operator- is the prefix set in config/default/kustomization.yaml file.
// Need to modify this name if prefix is changed in yaml file
aeroClusterServiceAccountName string = "aerospike-operator-controller-manager"
// This storage path annotation is added in pvc to make reverse association with storage.volume.path
// while deleting pvc
storageVolumeAnnotationKey = "storage-volume"
storageVolumeLegacyAnnotationKey = "storage-path"
confDirName = "confdir"
initConfDirName = "initconfigs"
podServiceAccountMountPath = "/var/run/secrets/kubernetes.io/serviceaccount"
)
type PortInfo struct {
connectionType string
configParam string
exposedOnHost bool
}
var defaultContainerPorts = map[string]PortInfo{
asdbv1.ServicePortName: {
connectionType: "service",
configParam: "port",
exposedOnHost: true,
},
asdbv1.ServiceTLSPortName: {
connectionType: "service",
configParam: "tls-port",
exposedOnHost: true,
},
asdbv1.FabricPortName: {
connectionType: "fabric",
configParam: "port",
},
asdbv1.FabricTLSPortName: {
connectionType: "fabric",
configParam: "tls-port",
},
asdbv1.HeartbeatPortName: {
connectionType: "heartbeat",
configParam: "port",
},
asdbv1.HeartbeatTLSPortName: {
connectionType: "heartbeat",
configParam: "tls-port",
},
asdbv1.InfoPortName: {
connectionType: "info",
configParam: "port",
exposedOnHost: true,
},
}
func (r *SingleClusterReconciler) createSTS(
namespacedName types.NamespacedName, rackState *RackState,
) (*appsv1.StatefulSet, error) {
replicas := int32(rackState.Size)
r.Log.Info("Create statefulset for AerospikeCluster", "size", replicas)
ports := getSTSContainerPort(
r.aeroCluster.Spec.PodSpec.MultiPodPerHost,
r.aeroCluster.Spec.AerospikeConfig,
)
operatorDefinedLabels := utils.LabelsForAerospikeClusterRack(
r.aeroCluster.Name, rackState.Rack.ID,
)
tlsName, _ := asdbv1.GetServiceTLSNameAndPort(r.aeroCluster.Spec.AerospikeConfig)
envVarList := []corev1.EnvVar{
newSTSEnvVar("MY_POD_NAME", "metadata.name"),
newSTSEnvVar("MY_POD_NAMESPACE", "metadata.namespace"),
newSTSEnvVar("MY_POD_IP", "status.podIP"),
newSTSEnvVar("MY_HOST_IP", "status.hostIP"),
newSTSEnvVarStatic("MY_POD_TLS_NAME", tlsName),
newSTSEnvVarStatic("MY_POD_CLUSTER_NAME", r.aeroCluster.Name),
}
st := &appsv1.StatefulSet{
ObjectMeta: metav1.ObjectMeta{
Name: namespacedName.Name,
Namespace: namespacedName.Namespace,
Labels: operatorDefinedLabels,
},
Spec: appsv1.StatefulSetSpec{
PodManagementPolicy: appsv1.ParallelPodManagement,
UpdateStrategy: appsv1.StatefulSetUpdateStrategy{
Type: appsv1.OnDeleteStatefulSetStrategyType,
},
Replicas: &replicas,
Selector: &metav1.LabelSelector{
MatchLabels: operatorDefinedLabels,
},
ServiceName: r.aeroCluster.Name,
Template: corev1.PodTemplateSpec{
ObjectMeta: metav1.ObjectMeta{
Labels: operatorDefinedLabels,
},
Spec: corev1.PodSpec{
ServiceAccountName: aeroClusterServiceAccountName,
// TerminationGracePeriodSeconds: &int64(30),
InitContainers: []corev1.Container{
{
Name: asdbv1.AerospikeInitContainerName,
Image: asdbv1.GetAerospikeInitContainerImage(r.aeroCluster),
ImagePullPolicy: corev1.PullIfNotPresent,
VolumeMounts: getDefaultAerospikeInitContainerVolumeMounts(),
Env: append(
envVarList, []corev1.EnvVar{
{
// Headless service has the same name as AerospikeCluster
Name: "SERVICE",
Value: getSTSHeadLessSvcName(r.aeroCluster),
},
// TODO: Do we need this var?
{
Name: "CONFIG_MAP_NAME",
Value: utils.GetNamespacedNameForSTSOrConfigMap(
r.aeroCluster, rackState.Rack.ID,
).Name,
},
}...,
),
},
},
Containers: []corev1.Container{
{
Name: asdbv1.AerospikeServerContainerName,
Image: r.aeroCluster.Spec.Image,
ImagePullPolicy: corev1.PullIfNotPresent,
Ports: ports,
Env: envVarList,
VolumeMounts: getDefaultAerospikeContainerVolumeMounts(),
ReadinessProbe: r.getReadinessProbe(),
},
},
Volumes: getDefaultSTSVolumes(r.aeroCluster, rackState),
},
},
},
}
r.updateSTSFromPodSpec(st, rackState)
// TODO: Add validation. device, file, both should not exist in same storage class
r.updateSTSStorage(st, rackState)
// Set AerospikeCluster instance as the owner and controller
if err := controllerutil.SetControllerReference(
r.aeroCluster, st, r.Scheme,
); err != nil {
return nil, err
}
if err := r.Client.Create(context.TODO(), st, createOption); err != nil {
return nil, fmt.Errorf("failed to create new StatefulSet: %v", err)
}
r.Log.Info(
"Created new StatefulSet", "StatefulSet.Namespace", st.Namespace,
"StatefulSet.Name", st.Name,
)
if err := r.waitForSTSToBeReady(st, nil); err != nil {
return st, fmt.Errorf(
"failed to wait for statefulset to be ready: %v", err,
)
}
return r.getSTS(rackState)
}
func (r *SingleClusterReconciler) getReadinessProbe() *corev1.Probe {
var readinessPort *int
if _, tlsPort := asdbv1.GetTLSNameAndPort(r.aeroCluster.Spec.AerospikeConfig, asdbv1.ServicePortName); tlsPort != nil {
readinessPort = tlsPort
} else {
readinessPort = asdbv1.GetServicePort(r.aeroCluster.Spec.AerospikeConfig)
}
return &corev1.Probe{
ProbeHandler: corev1.ProbeHandler{
TCPSocket: &corev1.TCPSocketAction{
Port: intstr.IntOrString{
Type: intstr.Int,
IntVal: int32(*readinessPort),
},
},
},
InitialDelaySeconds: 2,
PeriodSeconds: 5,
}
}
func (r *SingleClusterReconciler) isReadinessPortUpdated(pod *corev1.Pod) bool {
for idx := range pod.Spec.Containers {
container := &pod.Spec.Containers[idx]
if container.Name != asdbv1.AerospikeServerContainerName {
continue
}
// ignore if readiness probe is not set. Avoid rolling restart for old versions of operator
if container.ReadinessProbe == nil {
return false
}
if container.ReadinessProbe.TCPSocket != nil &&
container.ReadinessProbe.TCPSocket.String() != r.getReadinessProbe().TCPSocket.String() {
return true
}
}
return false
}
func (r *SingleClusterReconciler) deleteSTS(st *appsv1.StatefulSet) error {
r.Log.Info("Delete statefulset")
// No need to do cleanup pods after deleting sts
// It is only deleted while its creation is failed
// While doing rackRemove, we call scaleDown to 0 so that will do cleanup
return r.Client.Delete(context.TODO(), st)
}
func (r *SingleClusterReconciler) waitForSTSToBeReady(
st *appsv1.StatefulSet, ignorablePodNames sets.Set[string],
) error {
const (
podStatusMaxRetry = 18
podStatusRetryInterval = time.Second * 10
)
r.Log.Info(
"Waiting for statefulset to be ready", "WaitTimePerPod",
podStatusRetryInterval*time.Duration(podStatusMaxRetry),
)
var podIndex int32
for podIndex = 0; podIndex < *st.Spec.Replicas; podIndex++ {
podName := getSTSPodName(st.Name, podIndex)
if ignorablePodNames.Has(podName) {
continue
}
var isReady bool
pod := &corev1.Pod{}
// Wait for 10 sec to pod to get started
for i := 0; i < 5; i++ {
if err := r.Client.Get(
context.TODO(),
types.NamespacedName{Name: podName, Namespace: st.Namespace},
pod,
); err == nil {
break
}
time.Sleep(time.Second * 2)
}
// Wait for pod to get ready
for i := 0; i < podStatusMaxRetry; i++ {
r.Log.V(1).Info(
"Check statefulSet pod running and ready", "pod", podName,
)
if err := r.Client.Get(
context.TODO(),
types.NamespacedName{Name: podName, Namespace: st.Namespace},
pod,
); err != nil {
return fmt.Errorf(
"failed to get statefulSet pod %s: %v", podName, err,
)
}
if err := utils.CheckPodFailed(pod); err != nil {
return fmt.Errorf("statefulSet pod %s failed: %v", podName, err)
}
if utils.IsPodRunningAndReady(pod) {
isReady = true
r.Log.Info("Pod is running and ready", "pod", podName)
break
}
time.Sleep(podStatusRetryInterval)
}
if !isReady {
statusErr := fmt.Errorf(
"statefulSet pod is not ready. Status: %v",
pod.Status.Conditions,
)
r.Log.Error(statusErr, "Statefulset Not ready")
return statusErr
}
}
// Check for statefulset at the end,
// if we check before pods then we would not know status of individual pods
const (
stsStatusMaxRetry = 10
stsStatusRetryInterval = time.Second * 2
)
var updated bool
for i := 0; i < stsStatusMaxRetry; i++ {
time.Sleep(stsStatusRetryInterval)
r.Log.V(1).Info("Check statefulSet status is updated or not")
if err := r.Client.Get(
context.TODO(),
types.NamespacedName{Name: st.Name, Namespace: st.Namespace}, st,
); err != nil {
return err
}
if *st.Spec.Replicas == st.Status.Replicas {
updated = true
break
}
r.Log.V(1).Info(
"StatefulSet spec.replica not matching status.replica", "status",
st.Status.Replicas, "spec", *st.Spec.Replicas,
)
}
if !updated {
return fmt.Errorf("statefulset status is not updated")
}
r.Log.Info("StatefulSet is ready")
return nil
}
func (r *SingleClusterReconciler) getSTS(rackState *RackState) (*appsv1.StatefulSet, error) {
found := &appsv1.StatefulSet{}
if err := r.Client.Get(
context.TODO(),
utils.GetNamespacedNameForSTSOrConfigMap(r.aeroCluster, rackState.Rack.ID),
found,
); err != nil {
return nil, err
}
return found, nil
}
func (r *SingleClusterReconciler) buildSTSConfigMap(
namespacedName types.NamespacedName, rack *asdbv1.Rack,
) error {
r.Log.Info("Creating a new ConfigMap for statefulSet")
confMap := &corev1.ConfigMap{}
err := r.Client.Get(
context.TODO(), types.NamespacedName{
Name: namespacedName.Name, Namespace: namespacedName.Namespace,
}, confMap,
)
if err != nil {
if errors.IsNotFound(err) {
// build the aerospike config file based on the current spec
var configMapData map[string]string
configMapData, err = r.createConfigMapData(rack)
if err != nil {
return fmt.Errorf("failed to build dotConfig from map: %v", err)
}
ls := utils.LabelsForAerospikeCluster(r.aeroCluster.Name)
// return a configmap object containing aerospikeConfig
confMap = &corev1.ConfigMap{
ObjectMeta: metav1.ObjectMeta{
Name: namespacedName.Name,
Labels: ls,
Namespace: namespacedName.Namespace,
},
Data: configMapData,
}
// Set AerospikeCluster instance as the owner and controller
err = controllerutil.SetControllerReference(
r.aeroCluster, confMap, r.Scheme,
)
if err != nil {
return err
}
if err = r.Client.Create(
context.TODO(), confMap, createOption,
); err != nil {
return fmt.Errorf(
"failed to create new confMap for StatefulSet: %v", err,
)
}
r.Log.Info(
"Created new ConfigMap", "ConfigMap.Namespace",
confMap.Namespace, "ConfigMap.Name", confMap.Name,
)
return nil
}
return err
}
r.Log.Info(
"Configmap already exists for statefulSet - using existing configmap",
"name", utils.NamespacedName(confMap.Namespace, confMap.Name),
)
// Update existing configmap as it might not be current.
configMapData, err := r.createConfigMapData(rack)
if err != nil {
return fmt.Errorf("failed to build config map data: %v", err)
}
// Replace config map data if differs since we are supposed to create a new config map.
if reflect.DeepEqual(confMap.Data, configMapData) {
return nil
}
r.Log.Info(
"Updating existed configmap",
"name", utils.NamespacedName(confMap.Namespace, confMap.Name),
)
confMap.Data = configMapData
if err := r.Client.Update(
context.TODO(), confMap, updateOption,
); err != nil {
return fmt.Errorf("failed to update ConfigMap for StatefulSet: %v", err)
}
return nil
}
func (r *SingleClusterReconciler) updateSTSConfigMap(
namespacedName types.NamespacedName, rack *asdbv1.Rack,
) error {
r.Log.Info("Updating ConfigMap", "ConfigMap", namespacedName)
confMap := &corev1.ConfigMap{}
if err := r.Client.Get(context.TODO(), namespacedName, confMap); err != nil {
return err
}
// build the aerospike config file based on the current spec
configMapData, err := r.createConfigMapData(rack)
if err != nil {
return fmt.Errorf("failed to build dotConfig from map: %v", err)
}
// Overwrite only spec based keys. Do not touch other keys like pod metadata.
for k, v := range configMapData {
confMap.Data[k] = v
}
if err := r.Client.Update(
context.TODO(), confMap, updateOption,
); err != nil {
return fmt.Errorf("failed to update confMap for StatefulSet: %v", err)
}
return nil
}
// isPodUpgraded checks if all containers in the pod have images from the
// cluster spec.
func (r *SingleClusterReconciler) isPodUpgraded(pod *corev1.Pod) bool {
if !utils.IsPodRunningAndReady(pod) {
return false
}
return r.isPodOnDesiredImage(pod, false)
}
// isPodOnDesiredImage indicates if the pod is ready and on desired images for
// all containers.
func (r *SingleClusterReconciler) isPodOnDesiredImage(
pod *corev1.Pod, logChanges bool,
) bool {
return r.allContainersAreOnDesiredImages(
r.aeroCluster,
pod.Name, pod.Spec.Containers, logChanges,
) &&
r.allContainersAreOnDesiredImages(
r.aeroCluster,
pod.Name, pod.Spec.InitContainers,
logChanges,
)
}
func (r *SingleClusterReconciler) allContainersAreOnDesiredImages(
aeroCluster *asdbv1.AerospikeCluster, podName string,
containers []corev1.Container, logChanges bool,
) bool {
for idx := range containers {
container := &containers[idx]
desiredImage, err := utils.GetDesiredImage(aeroCluster, container.Name)
if err != nil {
// Maybe a deleted sidecar. Ignore.
continue
}
// TODO: Should we check status here?
// status may not be ready due to any bad config (e.g. bad podSpec).
// Due to this check, flow will be stuck at this place (upgradeImage)
// status := getPodContainerStatus(pod, ps.Name)
// if status == nil || !status.Ready || !IsImageEqual(ps.Image, desiredImage) {
// return false
// }
if !utils.IsImageEqual(container.Image, desiredImage) {
if container.Name == asdbv1.AerospikeInitContainerName {
if logChanges {
r.Log.Info(
"Ignoring change to Aerospike Init Image", "pod",
podName, "container", container.Name, "currentImage",
container.Image, "desiredImage", desiredImage,
)
}
continue
}
if logChanges {
r.Log.Info(
"Found container for upgrading/downgrading in pod", "pod",
podName, "container", container.Name, "currentImage",
container.Image, "desiredImage", desiredImage,
)
}
return false
}
}
return true
}
// Called while creating new cluster and also during rolling restart.
// Note: STS podSpec must be updated before updating storage
func (r *SingleClusterReconciler) updateSTSStorage(
st *appsv1.StatefulSet, rackState *RackState,
) {
r.updateSTSPVStorage(st, rackState)
r.updateSTSNonPVStorage(st, rackState)
// Sort volume attachments so that overlapping paths do not shadow each other.
// For e.g. mount for /etc/ should be listed before mount for /etc/aerospike
// otherwise /etc/aerospike will get shadowed.
sortContainerVolumeAttachments(st.Spec.Template.Spec.InitContainers)
sortContainerVolumeAttachments(st.Spec.Template.Spec.Containers)
}
func (r *SingleClusterReconciler) updateSTSPorts(
st *appsv1.StatefulSet,
) {
ports := getSTSContainerPort(
r.aeroCluster.Spec.PodSpec.MultiPodPerHost,
r.aeroCluster.Spec.AerospikeConfig,
)
st.Spec.Template.Spec.Containers[0].Ports = ports
}
func sortContainerVolumeAttachments(containers []corev1.Container) {
for idx := range containers {
sort.Slice(
containers[idx].VolumeMounts, func(p, q int) bool {
return containers[idx].VolumeMounts[p].MountPath < containers[idx].VolumeMounts[q].MountPath
},
)
sort.Slice(
containers[idx].VolumeDevices, func(p, q int) bool {
return containers[idx].VolumeDevices[p].DevicePath < containers[idx].VolumeDevices[q].DevicePath
},
)
}
}
// updateSTS updates the statefulset to match the spec. It is idempotent.
func (r *SingleClusterReconciler) updateSTS(
statefulSet *appsv1.StatefulSet, rackState *RackState,
) error {
// Update settings from pod spec.
r.updateSTSFromPodSpec(statefulSet, rackState)
// Updating ports when switching between tls and non-tls.
r.updateSTSPorts(statefulSet)
// Update the images for all containers from the spec.
// Our Pod Spec does not contain image for the Aerospike Server
// Container.
r.updateContainerImages(statefulSet)
// Updates the readiness probe TCP Port if changed for the aerospike server container
r.updateReadinessProbe(statefulSet)
// This should be called before updating storage
r.initializeSTSStorage(statefulSet, rackState)
// TODO: Add validation. device, file, both should not exist in same storage class
r.updateSTSStorage(statefulSet, rackState)
if err := retry.RetryOnConflict(retry.DefaultRetry, func() error {
found, err := r.getSTS(rackState)
if err != nil {
return err
}
// Save the updated stateful set.
found.Spec = statefulSet.Spec
return r.Client.Update(context.TODO(), found, updateOption)
}); err != nil {
return fmt.Errorf(
"failed to update StatefulSet %s: %v",
statefulSet.Name,
err,
)
}
r.Log.V(1).Info(
"Saved StatefulSet", "statefulSet", *statefulSet,
)
return nil
}
// Returns external storage devices and the corresponding source
// that have been attached to the containers directly.
// Devices that are attached to containers and not present in aerospike config spec
// as well status are considered as external devices.
func (r *SingleClusterReconciler) getExternalStorageDevices(
container *corev1.Container, rackState *RackState, stSpecVolumes []corev1.Volume,
) ([]corev1.VolumeDevice, []corev1.Volume) {
var (
externalDevices []corev1.VolumeDevice
volumesForDevice []corev1.Volume
)
rackStatusVolumes := r.getRackStatusVolumes(rackState)
for _, volumeDevice := range container.VolumeDevices {
volumeInSpec := getStorageVolume(rackState.Rack.Storage.Volumes, volumeDevice.Name)
volumeInStatus := getStorageVolume(rackStatusVolumes, volumeDevice.Name)
if volumeInSpec == nil && volumeInStatus == nil {
externalDevices = append(externalDevices, volumeDevice)
for idx := range stSpecVolumes {
volume := stSpecVolumes[idx]
if volume.Name == volumeDevice.Name {
volumesForDevice = append(volumesForDevice, volume)
}
}
}
}
return externalDevices, volumesForDevice
}
// Returns external volume mounts and the corresponding source
// that have been mounted to the containers directly.
// Volumes that are mounted to containers and not present in aerospike config spec
// as well status and are not one of the default volume mounts
// are considered as external Volumes.
func (r *SingleClusterReconciler) getExternalStorageMounts(
container *corev1.Container, rackState *RackState, stSpecVolumes []corev1.Volume,
) ([]corev1.VolumeMount, []corev1.Volume) {
var (
externalMounts []corev1.VolumeMount
volumesForMounts []corev1.Volume
)
rackStatusVolumes := r.getRackStatusVolumes(rackState)
for _, volumeMount := range container.VolumeMounts {
volumeInSpec := getStorageVolume(rackState.Rack.Storage.Volumes, volumeMount.Name)
volumeInStatus := getStorageVolume(rackStatusVolumes, volumeMount.Name)
volumeInDefault := getContainerVolumeMounts(getDefaultAerospikeInitContainerVolumeMounts(), volumeMount.Name)
if volumeInSpec == nil && volumeInStatus == nil && volumeInDefault == nil {
externalMounts = append(externalMounts, volumeMount)
for idx := range stSpecVolumes {
volume := stSpecVolumes[idx]
if volume.Name == volumeMount.Name {
volumesForMounts = append(volumesForMounts, volume)
}
}
}
}
return externalMounts, volumesForMounts
}
func (r *SingleClusterReconciler) updateSTSPVStorage(
st *appsv1.StatefulSet, rackState *RackState,
) {
volumes := rackState.Rack.Storage.GetPVs()
for idx := range volumes {
volume := &volumes[idx]
initContainerAttachments, containerAttachments := getFinalVolumeAttachmentsForVolume(volume)
switch {
case volume.Source.PersistentVolume.VolumeMode == corev1.PersistentVolumeBlock:
initContainerVolumePathPrefix := "/workdir/block-volumes"
r.Log.V(1).Info("Added volume device for volume", "volume", volume)
addVolumeDeviceInContainer(
volume.Name, initContainerAttachments,
st.Spec.Template.Spec.InitContainers,
initContainerVolumePathPrefix,
)
addVolumeDeviceInContainer(
volume.Name, containerAttachments,
st.Spec.Template.Spec.Containers, "",
)
case volume.Source.PersistentVolume.VolumeMode == corev1.PersistentVolumeFilesystem:
initContainerVolumePathPrefix := "/workdir/filesystem-volumes"
r.Log.V(1).Info("Added volume mount for volume", "volume", volume)
addVolumeMountInContainer(
volume.Name, initContainerAttachments,
st.Spec.Template.Spec.InitContainers,
initContainerVolumePathPrefix,
)
addVolumeMountInContainer(
volume.Name, containerAttachments,
st.Spec.Template.Spec.Containers, "",
)
default:
// Should never come here
continue
}
pvc := createPVCForVolumeAttachment(r.aeroCluster, volume)
if !containsElement(st.Spec.VolumeClaimTemplates, &pvc) {
st.Spec.VolumeClaimTemplates = append(
st.Spec.VolumeClaimTemplates, pvc,
)
r.Log.V(1).Info("Added PVC for volume", "volume", volume)
}
}
}
func containsElement(claims []corev1.PersistentVolumeClaim, query *corev1.PersistentVolumeClaim) bool {
for idx := range claims {
if claims[idx].Name == query.Name {
return true
}
}
return false
}
func (r *SingleClusterReconciler) updateSTSNonPVStorage(
st *appsv1.StatefulSet, rackState *RackState,
) {
volumes := rackState.Rack.Storage.GetNonPVs()
initContainerVolumePathPrefix := "/workdir/filesystem-volumes"
for idx := range volumes {
volume := &volumes[idx]
initContainerAttachments, containerAttachments := getFinalVolumeAttachmentsForVolume(volume)
r.Log.V(1).Info(
"Added volume mount in statefulSet pod containers for volume",
"volume", volume,
)
// Add volumeMount in statefulSet pod containers for volume
addVolumeMountInContainer(
volume.Name, initContainerAttachments,
st.Spec.Template.Spec.InitContainers, initContainerVolumePathPrefix,
)
addVolumeMountInContainer(
volume.Name, containerAttachments, st.Spec.Template.Spec.Containers,
"",
)
// Add volume in statefulSet template
k8sVolume := createVolumeForVolumeAttachment(volume)
st.Spec.Template.Spec.Volumes = append(
st.Spec.Template.Spec.Volumes, k8sVolume,
)
}
}
func (r *SingleClusterReconciler) updateSTSSchedulingPolicy(
st *appsv1.StatefulSet, rackState *RackState,
) {
affinity := &corev1.Affinity{}
// Use rack affinity, if given
if rackState.Rack.PodSpec.Affinity != nil {
affinity = lib.DeepCopy(rackState.Rack.PodSpec.Affinity).(*corev1.Affinity)
} else if r.aeroCluster.Spec.PodSpec.Affinity != nil {
affinity = lib.DeepCopy(r.aeroCluster.Spec.PodSpec.Affinity).(*corev1.Affinity)
}
// Set our rules in PodAntiAffinity
// only enable in production, so it can be used in 1 node clusters while debugging (minikube)
if !asdbv1.GetBool(r.aeroCluster.Spec.PodSpec.MultiPodPerHost) {
if affinity.PodAntiAffinity == nil {
affinity.PodAntiAffinity = &corev1.PodAntiAffinity{}
}
antiAffinityLabels := utils.LabelsForPodAntiAffinity(r.aeroCluster.Name)
r.Log.Info("Adding pod affinity rules for statefulSet pod")
antiAffinity := &corev1.PodAntiAffinity{
RequiredDuringSchedulingIgnoredDuringExecution: []corev1.PodAffinityTerm{
{
LabelSelector: &metav1.LabelSelector{
MatchLabels: antiAffinityLabels,
},
TopologyKey: "kubernetes.io/hostname",
},
},
}
affinity.PodAntiAffinity.RequiredDuringSchedulingIgnoredDuringExecution = append(
affinity.PodAntiAffinity.RequiredDuringSchedulingIgnoredDuringExecution,
antiAffinity.RequiredDuringSchedulingIgnoredDuringExecution...,
)
}
// Set our rules in NodeAffinity
var matchExpressions []corev1.NodeSelectorRequirement
if rackState.Rack.Zone != "" {
matchExpressions = append(
matchExpressions, corev1.NodeSelectorRequirement{
Key: "topology.kubernetes.io/zone",
Operator: corev1.NodeSelectorOpIn,
Values: []string{rackState.Rack.Zone},
},
)
}
if rackState.Rack.Region != "" {
matchExpressions = append(
matchExpressions, corev1.NodeSelectorRequirement{
Key: "topology.kubernetes.io/region",
Operator: corev1.NodeSelectorOpIn,
Values: []string{rackState.Rack.Region},
},
)
}
if rackState.Rack.RackLabel != "" {
matchExpressions = append(
matchExpressions, corev1.NodeSelectorRequirement{
Key: "aerospike.com/rack-label",
Operator: corev1.NodeSelectorOpIn,
Values: []string{rackState.Rack.RackLabel},
},
)
}
if rackState.Rack.NodeName != "" {
matchExpressions = append(
matchExpressions, corev1.NodeSelectorRequirement{
Key: "kubernetes.io/hostname",
Operator: corev1.NodeSelectorOpIn,
Values: []string{rackState.Rack.NodeName},
},
)
}
if len(r.aeroCluster.Spec.K8sNodeBlockList) > 0 {
matchExpressions = append(
matchExpressions, corev1.NodeSelectorRequirement{
Key: "kubernetes.io/hostname",
Operator: corev1.NodeSelectorOpNotIn,
Values: r.aeroCluster.Spec.K8sNodeBlockList,
},
)
}
if len(matchExpressions) != 0 {
if affinity.NodeAffinity == nil {
affinity.NodeAffinity = &corev1.NodeAffinity{}
}
selector := affinity.NodeAffinity.RequiredDuringSchedulingIgnoredDuringExecution
if selector == nil || len(selector.NodeSelectorTerms) == 0 {
ns := &corev1.NodeSelector{
NodeSelectorTerms: []corev1.NodeSelectorTerm{
{
MatchExpressions: matchExpressions,
},
},
}
affinity.NodeAffinity.RequiredDuringSchedulingIgnoredDuringExecution = ns
} else {
for idx := range selector.NodeSelectorTerms {
selector.NodeSelectorTerms[idx].MatchExpressions = append(
selector.NodeSelectorTerms[idx].MatchExpressions,
matchExpressions...,
)
}
}
}
st.Spec.Template.Spec.Affinity = affinity
// Use rack nodeSelector, if given
if len(rackState.Rack.PodSpec.NodeSelector) != 0 {
st.Spec.Template.Spec.NodeSelector = rackState.Rack.PodSpec.NodeSelector
} else {
st.Spec.Template.Spec.NodeSelector = r.aeroCluster.Spec.PodSpec.NodeSelector
}
// Use rack tolerations, if given
if len(rackState.Rack.PodSpec.Tolerations) != 0 {
st.Spec.Template.Spec.Tolerations = rackState.Rack.PodSpec.Tolerations
} else {
st.Spec.Template.Spec.Tolerations = r.aeroCluster.Spec.PodSpec.Tolerations
}
}
// Called while creating new cluster and also during rolling restart.
func (r *SingleClusterReconciler) updateSTSFromPodSpec(
st *appsv1.StatefulSet, rackState *RackState,
) {
defaultLabels := utils.LabelsForAerospikeClusterRack(
r.aeroCluster.Name, rackState.Rack.ID,
)
r.updateSTSSchedulingPolicy(st, rackState)
userDefinedLabels := r.aeroCluster.Spec.PodSpec.AerospikeObjectMeta.Labels
mergedLabels := utils.MergeLabels(defaultLabels, userDefinedLabels)
st.Spec.Template.Spec.HostNetwork = r.aeroCluster.Spec.PodSpec.HostNetwork
st.Spec.Template.ObjectMeta.Labels = mergedLabels
st.Spec.Template.ObjectMeta.Annotations = r.aeroCluster.Spec.PodSpec.AerospikeObjectMeta.Annotations
st.Spec.Template.Spec.DNSPolicy = r.aeroCluster.Spec.PodSpec.DNSPolicy
st.Spec.Template.Spec.DNSConfig = r.aeroCluster.Spec.PodSpec.DNSConfig