-
-
Notifications
You must be signed in to change notification settings - Fork 160
/
node.go
1208 lines (994 loc) · 32.6 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
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
//go:build linux && cgo && !agent
package db
import (
"context"
"database/sql"
"fmt"
"net/http"
"slices"
"strconv"
"strings"
"time"
"github.com/lxc/incus/v6/internal/server/db/cluster"
"github.com/lxc/incus/v6/internal/server/db/operationtype"
"github.com/lxc/incus/v6/internal/server/db/query"
localUtil "github.com/lxc/incus/v6/internal/server/util"
"github.com/lxc/incus/v6/internal/version"
"github.com/lxc/incus/v6/shared/api"
"github.com/lxc/incus/v6/shared/osarch"
)
// ClusterRole represents the role of a member in a cluster.
type ClusterRole string
// ClusterRoleDatabase represents the database role in a cluster.
const ClusterRoleDatabase = ClusterRole("database")
// ClusterRoleDatabaseStandBy represents the database stand-by role in a cluster.
const ClusterRoleDatabaseStandBy = ClusterRole("database-standby")
// ClusterRoleDatabaseLeader represents the database leader role in a cluster.
const ClusterRoleDatabaseLeader = ClusterRole("database-leader")
// ClusterRoleEventHub represents a cluster member who operates as an event hub.
const ClusterRoleEventHub = ClusterRole("event-hub")
// ClusterRoleOVNChassis represents a cluster member who operates as an OVN chassis.
const ClusterRoleOVNChassis = ClusterRole("ovn-chassis")
// ClusterRoles maps role ids into human-readable names.
//
// Note: the database role is currently stored directly in the raft
// configuration which acts as single source of truth for it. This map should
// only contain Incus-specific cluster roles.
var ClusterRoles = map[int]ClusterRole{
1: ClusterRoleEventHub,
2: ClusterRoleOVNChassis,
}
// Numeric type codes identifying different cluster member states.
const (
ClusterMemberStateCreated = 0
ClusterMemberStatePending = 1
ClusterMemberStateEvacuated = 2
)
// NodeInfo holds information about a single member in a cluster.
type NodeInfo struct {
ID int64 // Stable node identifier
Name string // User-assigned name of the node
Address string // Network address of the node
Description string // Node description (optional)
Schema int // Schema version of the daemon running the member
APIExtensions int // Number of API extensions of the daemon running the member
Heartbeat time.Time // Timestamp of the last heartbeat
Roles []ClusterRole // List of cluster roles
Architecture int // Node architecture
State int // Node state
Config map[string]string // Configuration for the node
Groups []string // Cluster groups
}
// IsOffline returns true if the last successful heartbeat time of the node is
// older than the given threshold.
func (n NodeInfo) IsOffline(threshold time.Duration) bool {
return nodeIsOffline(threshold, n.Heartbeat)
}
// NodeInfoArgs provides information about the cluster environment for use with NodeInfo.ToAPI().
type NodeInfoArgs struct {
LeaderAddress string
FailureDomains map[uint64]string
MemberFailureDomains map[string]uint64
OfflineThreshold time.Duration
MaxMemberVersion [2]int
RaftNodes []RaftNode
}
// ToAPI returns an API entry.
func (n NodeInfo) ToAPI(ctx context.Context, tx *ClusterTx, args NodeInfoArgs) (*api.ClusterMember, error) {
var err error
var maxVersion [2]int
var failureDomain string
domainID := args.MemberFailureDomains[n.Address]
failureDomain = args.FailureDomains[domainID]
// From local database.
var raftNode *RaftNode
for _, node := range args.RaftNodes {
if node.Address == n.Address {
raftNode = &node
break
}
}
// Fill in the struct.
result := api.ClusterMember{}
result.Description = n.Description
result.ServerName = n.Name
result.URL = fmt.Sprintf("https://%s", n.Address)
result.Database = false
result.Config = n.Config
result.Roles = make([]string, 0, len(n.Roles))
for _, r := range n.Roles {
result.Roles = append(result.Roles, string(r))
}
result.Groups = n.Groups
// Check if member is the leader.
if args.LeaderAddress == n.Address {
result.Roles = append(result.Roles, string(ClusterRoleDatabaseLeader))
result.Database = true
}
if raftNode != nil && raftNode.Role == RaftVoter {
result.Roles = append(result.Roles, string(ClusterRoleDatabase))
result.Database = true
}
if raftNode != nil && raftNode.Role == RaftStandBy {
result.Roles = append(result.Roles, string(ClusterRoleDatabaseStandBy))
result.Database = true
}
result.Architecture, err = osarch.ArchitectureName(n.Architecture)
if err != nil {
return nil, err
}
result.FailureDomain = failureDomain
// Set state and message.
result.Status = "Online"
result.Message = "Fully operational"
if n.State == ClusterMemberStateEvacuated {
result.Status = "Evacuated"
result.Message = "Unavailable due to maintenance"
} else if n.IsOffline(args.OfflineThreshold) {
result.Status = "Offline"
result.Message = fmt.Sprintf("No heartbeat for %s (%s)", time.Since(n.Heartbeat), n.Heartbeat)
} else {
// Check if up to date.
n, err := localUtil.CompareVersions(maxVersion, n.Version())
if err != nil {
return nil, err
}
if n == 1 {
result.Status = "Blocked"
result.Message = "Needs updating to newer version"
}
}
return &result, nil
}
// Version returns the node's version, composed by its schema level and
// number of extensions.
func (n NodeInfo) Version() [2]int {
return [2]int{n.Schema, n.APIExtensions}
}
// GetNodeByAddress returns the node with the given network address.
func (c *ClusterTx) GetNodeByAddress(ctx context.Context, address string) (NodeInfo, error) {
null := NodeInfo{}
nodes, err := c.nodes(ctx, false /* not pending */, "address=?", address)
if err != nil {
return null, err
}
switch len(nodes) {
case 0:
return null, api.StatusErrorf(http.StatusNotFound, "Cluster member not found")
case 1:
return nodes[0], nil
default:
return null, fmt.Errorf("more than one node matches")
}
}
// GetNodeMaxVersion returns the highest schema and API versions possible on the cluster.
func (c *ClusterTx) GetNodeMaxVersion(ctx context.Context) ([2]int, error) {
version := [2]int{}
// Get the maximum DB schema.
var maxSchema int
row := c.tx.QueryRowContext(ctx, "SELECT MAX(schema) FROM nodes")
err := row.Scan(&maxSchema)
if err != nil {
return version, err
}
// Get the maximum API extension.
var maxAPI int
row = c.tx.QueryRowContext(ctx, "SELECT MAX(api_extensions) FROM nodes")
err = row.Scan(&maxAPI)
if err != nil {
return version, err
}
// Compute the combined version.
version = [2]int{maxSchema, maxAPI}
return version, nil
}
// GetNodeWithID returns the node with the given ID.
func (c *ClusterTx) GetNodeWithID(ctx context.Context, nodeID int) (NodeInfo, error) {
null := NodeInfo{}
nodes, err := c.nodes(ctx, false /* not pending */, "id=?", nodeID)
if err != nil {
return null, err
}
switch len(nodes) {
case 0:
return null, api.StatusErrorf(http.StatusNotFound, "Cluster member not found")
case 1:
return nodes[0], nil
default:
return null, fmt.Errorf("More than one cluster member matches")
}
}
// GetPendingNodeByAddress returns the pending node with the given network address.
func (c *ClusterTx) GetPendingNodeByAddress(ctx context.Context, address string) (NodeInfo, error) {
null := NodeInfo{}
nodes, err := c.nodes(ctx, true /*pending */, "address=?", address)
if err != nil {
return null, err
}
switch len(nodes) {
case 0:
return null, api.StatusErrorf(http.StatusNotFound, "Cluster member not found")
case 1:
return nodes[0], nil
default:
return null, fmt.Errorf("More than one cluster member matches")
}
}
// GetNodeByName returns the node with the given name.
func (c *ClusterTx) GetNodeByName(ctx context.Context, name string) (NodeInfo, error) {
null := NodeInfo{}
nodes, err := c.nodes(ctx, false /* not pending */, "name=?", name)
if err != nil {
return null, err
}
switch len(nodes) {
case 0:
return null, api.StatusErrorf(http.StatusNotFound, "Cluster member not found")
case 1:
return nodes[0], nil
default:
return null, fmt.Errorf("More than one cluster member matches")
}
}
// GetLocalNodeName returns the name of the node this method is invoked on.
// Usually you should not use this function directly but instead use the cached State.ServerName value.
func (c *ClusterTx) GetLocalNodeName(ctx context.Context) (string, error) {
stmt := "SELECT name FROM nodes WHERE id=?"
names, err := query.SelectStrings(ctx, c.tx, stmt, c.nodeID)
if err != nil {
return "", err
}
switch len(names) {
case 0:
return "", nil
case 1:
return names[0], nil
default:
return "", fmt.Errorf("inconsistency: non-unique node ID")
}
}
// GetLocalNodeAddress returns the address of the node this method is invoked on.
func (c *ClusterTx) GetLocalNodeAddress(ctx context.Context) (string, error) {
stmt := "SELECT address FROM nodes WHERE id=?"
addresses, err := query.SelectStrings(ctx, c.tx, stmt, c.nodeID)
if err != nil {
return "", err
}
switch len(addresses) {
case 0:
return "", nil
case 1:
return addresses[0], nil
default:
return "", fmt.Errorf("inconsistency: non-unique node ID")
}
}
// NodeIsOutdated returns true if there's some cluster node having an API or
// schema version greater than the node this method is invoked on.
func (c *ClusterTx) NodeIsOutdated(ctx context.Context) (bool, error) {
nodes, err := c.nodes(ctx, false /* not pending */, "")
if err != nil {
return false, fmt.Errorf("Failed to fetch nodes: %w", err)
}
// Figure our own version.
version := [2]int{}
for _, node := range nodes {
if node.ID == c.nodeID {
version = node.Version()
}
}
if version[0] == 0 || version[1] == 0 {
return false, fmt.Errorf("Inconsistency: local member not found")
}
// Check if any of the other nodes is greater than us.
for _, node := range nodes {
if node.ID == c.nodeID {
continue
}
n, err := localUtil.CompareVersions(node.Version(), version)
if err != nil {
return false, fmt.Errorf("Failed to compare with version of member %s: %w", node.Name, err)
}
if n == 1 {
// The other node's version is greater than ours.
return true, nil
}
}
return false, nil
}
// GetNodes returns all cluster members that are part of the cluster.
//
// If this server is not clustered, a list with a single member whose address is 0.0.0.0 is returned.
func (c *ClusterTx) GetNodes(ctx context.Context) ([]NodeInfo, error) {
return c.nodes(ctx, false /* not pending */, "")
}
// GetNodesCount returns the number of members in the cluster.
//
// Since there's always at least one node row, even when not-clustered, the
// return value is greater than zero.
func (c *ClusterTx) GetNodesCount(ctx context.Context) (int, error) {
count, err := query.Count(ctx, c.tx, "nodes", "")
if err != nil {
return 0, fmt.Errorf("failed to count existing nodes: %w", err)
}
return count, nil
}
// RenameNode changes the name of an existing node.
//
// Return an error if a node with the same name already exists.
func (c *ClusterTx) RenameNode(ctx context.Context, old string, new string) error {
count, err := query.Count(ctx, c.tx, "nodes", "name=?", new)
if err != nil {
return fmt.Errorf("failed to check existing nodes: %w", err)
}
if count != 0 {
return ErrAlreadyDefined
}
stmt := `UPDATE nodes SET name=? WHERE name=?`
result, err := c.tx.Exec(stmt, new, old)
if err != nil {
return fmt.Errorf("failed to update node name: %w", err)
}
n, err := result.RowsAffected()
if err != nil {
return fmt.Errorf("failed to get rows count: %w", err)
}
if n != 1 {
return fmt.Errorf("expected to update one row, not %d", n)
}
return nil
}
// SetDescription changes the description of the given node.
func (c *ClusterTx) SetDescription(id int64, description string) error {
stmt := `UPDATE nodes SET description=? WHERE id=?`
result, err := c.tx.Exec(stmt, description, id)
if err != nil {
return fmt.Errorf("Failed to update node name: %w", err)
}
n, err := result.RowsAffected()
if err != nil {
return fmt.Errorf("Failed to get rows count: %w", err)
}
if n != 1 {
return fmt.Errorf("Expected to update one row, not %d", n)
}
return nil
}
// Nodes returns all members part of the cluster.
func (c *ClusterTx) nodes(ctx context.Context, pending bool, where string, args ...any) ([]NodeInfo, error) {
// Get node roles
sql := "SELECT node_id, role FROM nodes_roles"
nodeRoles := map[int64][]ClusterRole{}
err := query.Scan(ctx, c.Tx(), sql, func(scan func(dest ...any) error) error {
var nodeID int64
var role int
err := scan(&nodeID, &role)
if err != nil {
return err
}
if nodeRoles[nodeID] == nil {
nodeRoles[nodeID] = []ClusterRole{}
}
roleName := string(ClusterRoles[role])
nodeRoles[nodeID] = append(nodeRoles[nodeID], ClusterRole(roleName))
return nil
})
if err != nil && err.Error() != "no such table: nodes_roles" {
// Don't fail on a missing table, we need to handle updates
return nil, err
}
// Get node groups
sql = `SELECT node_id, cluster_groups.name FROM nodes_cluster_groups
JOIN cluster_groups ON cluster_groups.id = nodes_cluster_groups.group_id`
nodeGroups := map[int64][]string{}
err = query.Scan(ctx, c.Tx(), sql, func(scan func(dest ...any) error) error {
var nodeID int64
var group string
err := scan(&nodeID, &group)
if err != nil {
return err
}
if nodeGroups[nodeID] == nil {
nodeGroups[nodeID] = []string{}
}
nodeGroups[nodeID] = append(nodeGroups[nodeID], group)
return nil
})
if err != nil && err.Error() != "no such table: nodes_cluster_groups" {
// Don't fail on a missing table, we need to handle updates
return nil, err
}
// Get the node entries
sql = "SELECT id, name, address, description, schema, api_extensions, heartbeat, arch, state FROM nodes "
if pending {
// Include only pending nodes
sql += "WHERE state=? "
} else {
// Include created and evacuated nodes
sql += "WHERE state!=? "
}
args = append([]any{ClusterMemberStatePending}, args...)
if where != "" {
sql += fmt.Sprintf("AND %s ", where)
}
sql += "ORDER BY id"
// Process node entries
nodes := []NodeInfo{}
err = query.Scan(ctx, c.tx, sql, func(scan func(dest ...any) error) error {
node := NodeInfo{}
err := scan(&node.ID, &node.Name, &node.Address, &node.Description, &node.Schema, &node.APIExtensions, &node.Heartbeat, &node.Architecture, &node.State)
if err != nil {
return err
}
nodes = append(nodes, node)
return nil
}, args...)
if err != nil {
return nil, fmt.Errorf("Failed to fetch nodes: %w", err)
}
// Add the roles
for i, node := range nodes {
roles, ok := nodeRoles[node.ID]
if ok {
nodes[i].Roles = roles
}
}
// Add the groups
for i, node := range nodes {
groups, ok := nodeGroups[node.ID]
if ok {
nodes[i].Groups = groups
}
}
config, err := cluster.GetConfig(context.TODO(), c.Tx(), "node")
if err != nil {
return nil, fmt.Errorf("Failed to fetch nodes config: %w", err)
}
for i := range nodes {
data, ok := config[int(nodes[i].ID)]
if !ok {
nodes[i].Config = map[string]string{}
} else {
nodes[i].Config = data
}
}
return nodes, nil
}
// CreateNode adds a node to the current list of members that are part of the
// cluster. The node's architecture will be the architecture of the machine the
// method is being run on. It returns the ID of the newly inserted row.
func (c *ClusterTx) CreateNode(name string, address string) (int64, error) {
arch, err := osarch.ArchitectureGetLocalID()
if err != nil {
return -1, err
}
return c.CreateNodeWithArch(name, address, arch)
}
// CreateNodeWithArch is the same as NodeAdd, but lets setting the node
// architecture explicitly.
func (c *ClusterTx) CreateNodeWithArch(name string, address string, arch int) (int64, error) {
columns := []string{"name", "address", "schema", "api_extensions", "arch", "description"}
values := []any{name, address, cluster.SchemaVersion, version.APIExtensionsCount(), arch, ""}
return query.UpsertObject(c.tx, "nodes", columns, values)
}
// SetNodePendingFlag toggles the pending flag for the node. A node is pending when
// it's been accepted in the cluster, but has not yet actually joined it.
func (c *ClusterTx) SetNodePendingFlag(id int64, pending bool) error {
value := 0
if pending {
value = 1
}
result, err := c.tx.Exec("UPDATE nodes SET state=? WHERE id=?", value, id)
if err != nil {
return err
}
n, err := result.RowsAffected()
if err != nil {
return err
}
if n != 1 {
return fmt.Errorf("query updated %d rows instead of 1", n)
}
return nil
}
// BootstrapNode sets the name and address of the first cluster member, with id: 1.
func (c *ClusterTx) BootstrapNode(name string, address string) error {
result, err := c.tx.Exec("UPDATE nodes SET name=?, address=? WHERE id=1", name, address)
if err != nil {
return err
}
n, err := result.RowsAffected()
if err != nil {
return err
}
if n != 1 {
return fmt.Errorf("query updated %d rows instead of 1", n)
}
return nil
}
// UpdateNodeConfig updates the replaces the node's config with the specified config.
func (c *ClusterTx) UpdateNodeConfig(ctx context.Context, id int64, config map[string]string) error {
err := cluster.UpdateConfig(ctx, c.Tx(), "node", int(id), config)
if err != nil {
return fmt.Errorf("Unable to update node config: %w", err)
}
return nil
}
// UpdateNodeRoles changes the list of roles on a member.
func (c *ClusterTx) UpdateNodeRoles(id int64, roles []ClusterRole) error {
getRoleID := func(role ClusterRole) (int, error) {
for k, v := range ClusterRoles {
if v == role {
return k, nil
}
}
return -1, fmt.Errorf("Invalid cluster role %q", role)
}
// Translate role names to ids
roleIDs := []int{}
for _, role := range roles {
// Skip internal-only roles.
if role == ClusterRoleDatabase || role == ClusterRoleDatabaseStandBy || role == ClusterRoleDatabaseLeader {
continue
}
roleID, err := getRoleID(role)
if err != nil {
return err
}
roleIDs = append(roleIDs, roleID)
}
// Update the database record
_, err := c.tx.Exec("DELETE FROM nodes_roles WHERE node_id=?", id)
if err != nil {
return err
}
for _, roleID := range roleIDs {
_, err := c.tx.Exec("INSERT INTO nodes_roles (node_id, role) VALUES (?, ?)", id, roleID)
if err != nil {
return err
}
}
return nil
}
// UpdateNodeClusterGroups changes the list of cluster groups the member belongs to.
func (c *ClusterTx) UpdateNodeClusterGroups(ctx context.Context, id int64, groups []string) error {
nodeInfo, err := c.GetNodeWithID(ctx, int(id))
if err != nil {
return err
}
oldGroups, err := c.GetClusterGroupsWithNode(ctx, nodeInfo.Name)
if err != nil {
return err
}
skipGroups := []string{}
// Check if node already belongs to the given groups.
for _, newGroup := range groups {
if slices.Contains(oldGroups, newGroup) {
// Node already belongs to this group.
skipGroups = append(skipGroups, newGroup)
continue
}
// Add node to new group.
err = c.AddNodeToClusterGroup(ctx, newGroup, nodeInfo.Name)
if err != nil {
return fmt.Errorf("Failed to add member to cluster group: %w", err)
}
}
for _, oldGroup := range oldGroups {
if slices.Contains(skipGroups, oldGroup) {
continue
}
// Remove node from group.
err = c.RemoveNodeFromClusterGroup(ctx, oldGroup, nodeInfo.Name)
if err != nil {
return fmt.Errorf("Failed to remove member from cluster group: %w", err)
}
}
return nil
}
// UpdateNodeFailureDomain changes the failure domain of a node.
func (c *ClusterTx) UpdateNodeFailureDomain(ctx context.Context, id int64, domain string) error {
var domainID any
if domain == "" {
return fmt.Errorf("Failure domain name can't be empty")
}
if domain == "default" {
domainID = nil
} else {
row := c.tx.QueryRowContext(ctx, "SELECT id FROM nodes_failure_domains WHERE name=?", domain)
err := row.Scan(&domainID)
if err != nil {
if err != sql.ErrNoRows {
return fmt.Errorf("Load failure domain name: %w", err)
}
result, err := c.tx.Exec("INSERT INTO nodes_failure_domains (name) VALUES (?)", domain)
if err != nil {
return fmt.Errorf("Create new failure domain: %w", err)
}
domainID, err = result.LastInsertId()
if err != nil {
return fmt.Errorf("Get last inserted ID: %w", err)
}
}
}
result, err := c.tx.Exec("UPDATE nodes SET failure_domain_id=? WHERE id=?", domainID, id)
if err != nil {
return err
}
n, err := result.RowsAffected()
if err != nil {
return err
}
if n != 1 {
return fmt.Errorf("Query updated %d rows instead of 1", n)
}
return nil
}
// UpdateNodeStatus changes the state of a node.
func (c *ClusterTx) UpdateNodeStatus(id int64, state int) error {
result, err := c.tx.Exec("UPDATE nodes SET state=? WHERE id=?", state, id)
if err != nil {
return err
}
n, err := result.RowsAffected()
if err != nil {
return err
}
if n != 1 {
return fmt.Errorf("Query updated %d rows instead of 1", n)
}
return nil
}
// GetNodeFailureDomain returns the failure domain associated with the node with the given ID.
func (c *ClusterTx) GetNodeFailureDomain(ctx context.Context, id int64) (string, error) {
stmt := `
SELECT coalesce(nodes_failure_domains.name,'default')
FROM nodes LEFT JOIN nodes_failure_domains ON nodes.failure_domain_id = nodes_failure_domains.id
WHERE nodes.id=?
`
var domain string
err := c.tx.QueryRowContext(ctx, stmt, id).Scan(&domain)
if err != nil {
return "", err
}
return domain, nil
}
// GetNodesFailureDomains returns a map associating each node address with its
// failure domain code.
func (c *ClusterTx) GetNodesFailureDomains(ctx context.Context) (map[string]uint64, error) {
sql := "SELECT address, coalesce(failure_domain_id, 0) FROM nodes"
type failureDomain struct {
Address string
FailureDomainID int64
}
rows := []failureDomain{}
err := query.Scan(ctx, c.tx, sql, func(scan func(dest ...any) error) error {
fd := failureDomain{}
err := scan(&fd.Address, &fd.FailureDomainID)
if err != nil {
return err
}
rows = append(rows, fd)
return nil
})
if err != nil {
return nil, err
}
domains := map[string]uint64{}
for _, row := range rows {
domains[row.Address] = uint64(row.FailureDomainID)
}
return domains, nil
}
// GetFailureDomainsNames return a map associating failure domain IDs to their
// names.
func (c *ClusterTx) GetFailureDomainsNames(ctx context.Context) (map[uint64]string, error) {
sql := "SELECT id, name FROM nodes_failure_domains"
type failureDomain struct {
ID int64
Name string
}
rows := []failureDomain{}
err := query.Scan(ctx, c.tx, sql, func(scan func(dest ...any) error) error {
fd := failureDomain{}
err := scan(&fd.ID, &fd.Name)
if err != nil {
return err
}
rows = append(rows, fd)
return nil
})
if err != nil {
return nil, err
}
domains := map[uint64]string{
0: "default", // Default failure domain, when not set
}
for _, row := range rows {
domains[uint64(row.ID)] = row.Name
}
return domains, nil
}
// RemoveNode removes the node with the given id.
func (c *ClusterTx) RemoveNode(id int64) error {
result, err := c.tx.Exec("DELETE FROM nodes WHERE id=?", id)
if err != nil {
return err
}
n, err := result.RowsAffected()
if err != nil {
return err
}
if n != 1 {
return fmt.Errorf("query deleted %d rows instead of 1", n)
}
return nil
}
// SetNodeHeartbeat updates the heartbeat column of the node with the given address.
func (c *ClusterTx) SetNodeHeartbeat(address string, heartbeat time.Time) error {
stmt := "UPDATE nodes SET heartbeat=? WHERE address=?"
result, err := c.tx.Exec(stmt, heartbeat, address)
if err != nil {
return err
}
n, err := result.RowsAffected()
if err != nil {
return err
}
if n < 1 {
return api.StatusErrorf(http.StatusNotFound, "Cluster member not found")
} else if n > 1 {
return fmt.Errorf("Expected to update one row and not %d", n)
}
return nil
}
// NodeIsEmpty returns an empty string if the node with the given ID has no
// instances or images associated with it. Otherwise, it returns a message
// say what's left.
func (c *ClusterTx) NodeIsEmpty(ctx context.Context, id int64) (string, error) {
// Check if the node has any instances.
instances, err := query.SelectStrings(ctx, c.tx, "SELECT name FROM instances WHERE node_id=?", id)
if err != nil {
return "", fmt.Errorf("Failed to get instances for node %d: %w", id, err)
}
if len(instances) > 0 {
message := fmt.Sprintf(
"Node still has the following instances: %s", strings.Join(instances, ", "))
return message, nil
}
// Check if the node has any images available only in it.
type image struct {
fingerprint string
nodeID int64
}
images := []image{}
sql := `SELECT fingerprint, node_id FROM images JOIN images_nodes ON images.id=images_nodes.image_id`
err = query.Scan(ctx, c.tx, sql, func(scan func(dest ...any) error) error {
img := image{}
err := scan(&img.fingerprint, &img.nodeID)
if err != nil {
return err
}
images = append(images, img)
return nil
})
if err != nil {
return "", fmt.Errorf("Failed to get image list for node %d: %w", id, err)
}
index := map[string][]int64{} // Map fingerprints to IDs of nodes
for _, image := range images {
index[image.fingerprint] = append(index[image.fingerprint], image.nodeID)
}
fingerprints := []string{}
for fingerprint, ids := range index {
if len(ids) > 1 {
continue
}
if ids[0] == id {
fingerprints = append(fingerprints, fingerprint)
}
}
if len(fingerprints) > 0 {
message := fmt.Sprintf(
"Node still has the following images: %s", strings.Join(fingerprints, ", "))
return message, nil
}
// Check if the node has any custom volumes.
drivers := make([]string, len(StorageRemoteDriverNames()))
for i, entry := range StorageRemoteDriverNames() {
drivers[i] = fmt.Sprintf("'%s'", entry)
}
sql = `
SELECT storage_volumes.name
FROM storage_volumes
JOIN storage_pools ON storage_volumes.storage_pool_id=storage_pools.id
WHERE storage_volumes.node_id=? AND storage_volumes.type=? AND storage_pools.driver NOT IN (%s)
`
volumes, err := query.SelectStrings(ctx, c.tx, fmt.Sprintf(sql, strings.Join(drivers, ", ")), id, StoragePoolVolumeTypeCustom)
if err != nil {
return "", fmt.Errorf("Failed to get custom volumes for node %d: %w", id, err)
}
if len(volumes) > 0 {
message := fmt.Sprintf(
"Node still has the following custom volumes: %s", strings.Join(volumes, ", "))
return message, nil
}
return "", nil
}
// ClearNode removes any instance or image associated with this node.
func (c *ClusterTx) ClearNode(ctx context.Context, id int64) error {
_, err := c.tx.Exec("DELETE FROM instances WHERE node_id=?", id)
if err != nil {
return err
}
// Get the IDs of the images this node is hosting.