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
|
package topology
import (
"errors"
"fmt"
"math/rand"
"sync"
"time"
"github.com/chrislusf/seaweedfs/weed/glog"
"github.com/chrislusf/seaweedfs/weed/storage"
)
// mapping from volume to its locations, inverted from server to volume
type VolumeLayout struct {
rp *storage.ReplicaPlacement
ttl *storage.TTL
vid2location map[storage.VolumeId]*VolumeLocationList
writables []storage.VolumeId // transient array of writable volume id
readonlyVolumes map[storage.VolumeId]bool // transient set of readonly volumes
oversizedVolumes map[storage.VolumeId]bool // set of oversized volumes
volumeSizeLimit uint64
accessLock sync.RWMutex
}
type VolumeLayoutStats struct {
TotalSize uint64
UsedSize uint64
FileCount uint64
}
func NewVolumeLayout(rp *storage.ReplicaPlacement, ttl *storage.TTL, volumeSizeLimit uint64) *VolumeLayout {
return &VolumeLayout{
rp: rp,
ttl: ttl,
vid2location: make(map[storage.VolumeId]*VolumeLocationList),
writables: *new([]storage.VolumeId),
readonlyVolumes: make(map[storage.VolumeId]bool),
oversizedVolumes: make(map[storage.VolumeId]bool),
volumeSizeLimit: volumeSizeLimit,
}
}
func (vl *VolumeLayout) String() string {
return fmt.Sprintf("rp:%v, ttl:%v, vid2location:%v, writables:%v, volumeSizeLimit:%v", vl.rp, vl.ttl, vl.vid2location, vl.writables, vl.volumeSizeLimit)
}
func (vl *VolumeLayout) RegisterVolume(v *storage.VolumeInfo, dn *DataNode) {
vl.accessLock.Lock()
defer vl.accessLock.Unlock()
if _, ok := vl.vid2location[v.Id]; !ok {
vl.vid2location[v.Id] = NewVolumeLocationList()
}
vl.vid2location[v.Id].Set(dn)
// glog.V(4).Infof("volume %d added to %s len %d copy %d", v.Id, dn.Id(), vl.vid2location[v.Id].Length(), v.ReplicaPlacement.GetCopyCount())
for _, dn := range vl.vid2location[v.Id].list {
if vInfo, err := dn.GetVolumesById(v.Id); err == nil {
if vInfo.ReadOnly {
glog.V(3).Infof("vid %d removed from writable", v.Id)
vl.removeFromWritable(v.Id)
vl.readonlyVolumes[v.Id] = true
return
} else {
delete(vl.readonlyVolumes, v.Id)
}
} else {
glog.V(3).Infof("vid %d removed from writable", v.Id)
vl.removeFromWritable(v.Id)
delete(vl.readonlyVolumes, v.Id)
return
}
}
if vl.vid2location[v.Id].Length() == vl.rp.GetCopyCount() && vl.isWritable(v) {
if _, ok := vl.oversizedVolumes[v.Id]; !ok {
vl.addToWritable(v.Id)
}
} else {
vl.rememberOversizedVolumne(v)
vl.removeFromWritable(v.Id)
}
}
func (vl *VolumeLayout) rememberOversizedVolumne(v *storage.VolumeInfo) {
if vl.isOversized(v) {
vl.oversizedVolumes[v.Id] = true
}
}
func (vl *VolumeLayout) UnRegisterVolume(v *storage.VolumeInfo, dn *DataNode) {
vl.accessLock.Lock()
defer vl.accessLock.Unlock()
vl.removeFromWritable(v.Id)
delete(vl.vid2location, v.Id)
}
func (vl *VolumeLayout) addToWritable(vid storage.VolumeId) {
for _, id := range vl.writables {
if vid == id {
return
}
}
vl.writables = append(vl.writables, vid)
}
func (vl *VolumeLayout) isOversized(v *storage.VolumeInfo) bool {
return uint64(v.Size) >= vl.volumeSizeLimit
}
func (vl *VolumeLayout) isWritable(v *storage.VolumeInfo) bool {
return !vl.isOversized(v) &&
v.Version == storage.CurrentVersion &&
!v.ReadOnly
}
func (vl *VolumeLayout) isEmpty() bool {
vl.accessLock.RLock()
defer vl.accessLock.RUnlock()
return len(vl.vid2location) == 0
}
func (vl *VolumeLayout) Lookup(vid storage.VolumeId) []*DataNode {
vl.accessLock.RLock()
defer vl.accessLock.RUnlock()
if location := vl.vid2location[vid]; location != nil {
return location.list
}
return nil
}
func (vl *VolumeLayout) ListVolumeServers() (nodes []*DataNode) {
vl.accessLock.RLock()
defer vl.accessLock.RUnlock()
for _, location := range vl.vid2location {
nodes = append(nodes, location.list...)
}
return
}
func (vl *VolumeLayout) PickForWrite(count uint64, option *VolumeGrowOption) (*storage.VolumeId, uint64, *VolumeLocationList, error) {
vl.accessLock.RLock()
defer vl.accessLock.RUnlock()
lenWriters := len(vl.writables)
if lenWriters <= 0 {
glog.V(0).Infoln("No more writable volumes!")
return nil, 0, nil, errors.New("No more writable volumes!")
}
if option.DataCenter == "" {
vid := vl.writables[rand.Intn(lenWriters)]
locationList := vl.vid2location[vid]
if locationList != nil {
return &vid, count, locationList, nil
}
return nil, 0, nil, errors.New("Strangely vid " + vid.String() + " is on no machine!")
}
var vid storage.VolumeId
var locationList *VolumeLocationList
counter := 0
for _, v := range vl.writables {
volumeLocationList := vl.vid2location[v]
for _, dn := range volumeLocationList.list {
if dn.GetDataCenter().Id() == NodeId(option.DataCenter) {
if option.Rack != "" && dn.GetRack().Id() != NodeId(option.Rack) {
continue
}
if option.DataNode != "" && dn.Id() != NodeId(option.DataNode) {
continue
}
counter++
if rand.Intn(counter) < 1 {
vid, locationList = v, volumeLocationList
}
}
}
}
return &vid, count, locationList, nil
}
func (vl *VolumeLayout) GetActiveVolumeCount(option *VolumeGrowOption) int {
vl.accessLock.RLock()
defer vl.accessLock.RUnlock()
if option.DataCenter == "" {
return len(vl.writables)
}
counter := 0
for _, v := range vl.writables {
for _, dn := range vl.vid2location[v].list {
if dn.GetDataCenter().Id() == NodeId(option.DataCenter) {
if option.Rack != "" && dn.GetRack().Id() != NodeId(option.Rack) {
continue
}
if option.DataNode != "" && dn.Id() != NodeId(option.DataNode) {
continue
}
counter++
}
}
}
return counter
}
func (vl *VolumeLayout) removeFromWritable(vid storage.VolumeId) bool {
toDeleteIndex := -1
for k, id := range vl.writables {
if id == vid {
toDeleteIndex = k
break
}
}
if toDeleteIndex >= 0 {
glog.V(0).Infoln("Volume", vid, "becomes unwritable")
vl.writables = append(vl.writables[0:toDeleteIndex], vl.writables[toDeleteIndex+1:]...)
return true
}
return false
}
func (vl *VolumeLayout) setVolumeWritable(vid storage.VolumeId) bool {
for _, v := range vl.writables {
if v == vid {
return false
}
}
glog.V(0).Infoln("Volume", vid, "becomes writable")
vl.writables = append(vl.writables, vid)
return true
}
func (vl *VolumeLayout) SetVolumeUnavailable(dn *DataNode, vid storage.VolumeId) bool {
vl.accessLock.Lock()
defer vl.accessLock.Unlock()
if location, ok := vl.vid2location[vid]; ok {
if location.Remove(dn) {
if location.Length() < vl.rp.GetCopyCount() {
glog.V(0).Infoln("Volume", vid, "has", location.Length(), "replica, less than required", vl.rp.GetCopyCount())
return vl.removeFromWritable(vid)
}
}
}
return false
}
func (vl *VolumeLayout) SetVolumeAvailable(dn *DataNode, vid storage.VolumeId) bool {
vl.accessLock.Lock()
defer vl.accessLock.Unlock()
vl.vid2location[vid].Set(dn)
if vl.vid2location[vid].Length() >= vl.rp.GetCopyCount() {
return vl.setVolumeWritable(vid)
}
return false
}
func (vl *VolumeLayout) SetVolumeCapacityFull(vid storage.VolumeId) bool {
vl.accessLock.Lock()
defer vl.accessLock.Unlock()
// glog.V(0).Infoln("Volume", vid, "reaches full capacity.")
return vl.removeFromWritable(vid)
}
func (vl *VolumeLayout) ToMap() map[string]interface{} {
m := make(map[string]interface{})
m["replication"] = vl.rp.String()
m["ttl"] = vl.ttl.String()
m["writables"] = vl.writables
//m["locations"] = vl.vid2location
return m
}
func (vl *VolumeLayout) Stats() *VolumeLayoutStats {
vl.accessLock.RLock()
defer vl.accessLock.RUnlock()
ret := &VolumeLayoutStats{}
freshThreshold := time.Now().Unix() - 60
for vid, vll := range vl.vid2location {
size, fileCount := vll.Stats(vid, freshThreshold)
ret.FileCount += uint64(fileCount)
ret.UsedSize += size
if vl.readonlyVolumes[vid] {
ret.TotalSize += size
} else {
ret.TotalSize += vl.volumeSizeLimit
}
}
return ret
}
|
