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
|
package filersink
import (
"context"
"fmt"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/wdclient"
"math"
"google.golang.org/grpc"
"github.com/seaweedfs/seaweedfs/weed/security"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/replication/sink"
"github.com/seaweedfs/seaweedfs/weed/replication/source"
"github.com/seaweedfs/seaweedfs/weed/util"
)
type FilerSink struct {
filerSource *source.FilerSource
grpcAddress string
dir string
replication string
collection string
ttlSec int32
diskType string
dataCenter string
grpcDialOption grpc.DialOption
address string
writeChunkByFiler bool
isIncremental bool
}
func init() {
sink.Sinks = append(sink.Sinks, &FilerSink{})
}
func (fs *FilerSink) GetName() string {
return "filer"
}
func (fs *FilerSink) GetSinkToDirectory() string {
return fs.dir
}
func (fs *FilerSink) IsIncremental() bool {
return fs.isIncremental
}
func (fs *FilerSink) Initialize(configuration util.Configuration, prefix string) error {
fs.isIncremental = configuration.GetBool(prefix + "is_incremental")
return fs.DoInitialize(
"",
configuration.GetString(prefix+"grpcAddress"),
configuration.GetString(prefix+"directory"),
configuration.GetString(prefix+"replication"),
configuration.GetString(prefix+"collection"),
configuration.GetInt(prefix+"ttlSec"),
configuration.GetString(prefix+"disk"),
security.LoadClientTLS(util.GetViper(), "grpc.client"),
false)
}
func (fs *FilerSink) SetSourceFiler(s *source.FilerSource) {
fs.filerSource = s
}
func (fs *FilerSink) DoInitialize(address, grpcAddress string, dir string,
replication string, collection string, ttlSec int, diskType string, grpcDialOption grpc.DialOption, writeChunkByFiler bool) (err error) {
fs.address = address
if fs.address == "" {
fs.address = pb.GrpcAddressToServerAddress(grpcAddress)
}
fs.grpcAddress = grpcAddress
fs.dir = dir
fs.replication = replication
fs.collection = collection
fs.ttlSec = int32(ttlSec)
fs.diskType = diskType
fs.grpcDialOption = grpcDialOption
fs.writeChunkByFiler = writeChunkByFiler
return nil
}
func (fs *FilerSink) DeleteEntry(key string, isDirectory, deleteIncludeChunks bool, signatures []int32) error {
dir, name := util.FullPath(key).DirAndName()
glog.V(4).Infof("delete entry: %v", key)
err := filer_pb.Remove(fs, dir, name, deleteIncludeChunks, true, true, true, signatures)
if err != nil {
glog.V(0).Infof("delete entry %s: %v", key, err)
return fmt.Errorf("delete entry %s: %v", key, err)
}
return nil
}
func (fs *FilerSink) CreateEntry(key string, entry *filer_pb.Entry, signatures []int32) error {
return fs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
dir, name := util.FullPath(key).DirAndName()
// look up existing entry
lookupRequest := &filer_pb.LookupDirectoryEntryRequest{
Directory: dir,
Name: name,
}
glog.V(1).Infof("lookup: %v", lookupRequest)
if resp, err := filer_pb.LookupEntry(client, lookupRequest); err == nil {
if filer.ETag(resp.Entry) == filer.ETag(entry) {
glog.V(3).Infof("already replicated %s", key)
return nil
}
}
replicatedChunks, err := fs.replicateChunks(entry.Chunks, key)
if err != nil {
// only warning here since the source chunk may have been deleted already
glog.Warningf("replicate entry chunks %s: %v", key, err)
}
glog.V(4).Infof("replicated %s %+v ===> %+v", key, entry.Chunks, replicatedChunks)
request := &filer_pb.CreateEntryRequest{
Directory: dir,
Entry: &filer_pb.Entry{
Name: name,
IsDirectory: entry.IsDirectory,
Attributes: entry.Attributes,
Chunks: replicatedChunks,
Content: entry.Content,
RemoteEntry: entry.RemoteEntry,
},
IsFromOtherCluster: true,
Signatures: signatures,
}
glog.V(3).Infof("create: %v", request)
if err := filer_pb.CreateEntry(client, request); err != nil {
glog.V(0).Infof("create entry %s: %v", key, err)
return fmt.Errorf("create entry %s: %v", key, err)
}
return nil
})
}
func (fs *FilerSink) UpdateEntry(key string, oldEntry *filer_pb.Entry, newParentPath string, newEntry *filer_pb.Entry, deleteIncludeChunks bool, signatures []int32) (foundExistingEntry bool, err error) {
dir, name := util.FullPath(key).DirAndName()
// read existing entry
var existingEntry *filer_pb.Entry
err = fs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
request := &filer_pb.LookupDirectoryEntryRequest{
Directory: dir,
Name: name,
}
glog.V(4).Infof("lookup entry: %v", request)
resp, err := filer_pb.LookupEntry(client, request)
if err != nil {
glog.V(0).Infof("lookup %s: %v", key, err)
return err
}
existingEntry = resp.Entry
return nil
})
if err != nil {
return false, fmt.Errorf("lookup %s: %v", key, err)
}
glog.V(4).Infof("oldEntry %+v, newEntry %+v, existingEntry: %+v", oldEntry, newEntry, existingEntry)
if existingEntry.Attributes.Mtime > newEntry.Attributes.Mtime {
// skip if already changed
// this usually happens when the messages are not ordered
glog.V(2).Infof("late updates %s", key)
} else if filer.ETag(newEntry) == filer.ETag(existingEntry) {
// skip if no change
// this usually happens when retrying the replication
glog.V(3).Infof("already replicated %s", key)
} else {
// find out what changed
deletedChunks, newChunks, err := compareChunks(filer.LookupFn(fs), oldEntry, newEntry)
if err != nil {
return true, fmt.Errorf("replicte %s compare chunks error: %v", key, err)
}
// delete the chunks that are deleted from the source
if deleteIncludeChunks {
// remove the deleted chunks. Actual data deletion happens in filer UpdateEntry FindUnusedFileChunks
existingEntry.Chunks = filer.DoMinusChunksBySourceFileId(existingEntry.Chunks, deletedChunks)
}
// replicate the chunks that are new in the source
replicatedChunks, err := fs.replicateChunks(newChunks, key)
if err != nil {
return true, fmt.Errorf("replicte %s chunks error: %v", key, err)
}
existingEntry.Chunks = append(existingEntry.Chunks, replicatedChunks...)
existingEntry.Attributes = newEntry.Attributes
existingEntry.Extended = newEntry.Extended
existingEntry.HardLinkId = newEntry.HardLinkId
existingEntry.HardLinkCounter = newEntry.HardLinkCounter
existingEntry.Content = newEntry.Content
existingEntry.RemoteEntry = newEntry.RemoteEntry
}
// save updated meta data
return true, fs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
request := &filer_pb.UpdateEntryRequest{
Directory: newParentPath,
Entry: existingEntry,
IsFromOtherCluster: true,
Signatures: signatures,
}
if _, err := client.UpdateEntry(context.Background(), request); err != nil {
return fmt.Errorf("update existingEntry %s: %v", key, err)
}
return nil
})
}
func compareChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, oldEntry, newEntry *filer_pb.Entry) (deletedChunks, newChunks []*filer_pb.FileChunk, err error) {
aData, aMeta, aErr := filer.ResolveChunkManifest(lookupFileIdFn, oldEntry.Chunks, 0, math.MaxInt64)
if aErr != nil {
return nil, nil, aErr
}
bData, bMeta, bErr := filer.ResolveChunkManifest(lookupFileIdFn, newEntry.Chunks, 0, math.MaxInt64)
if bErr != nil {
return nil, nil, bErr
}
deletedChunks = append(deletedChunks, filer.DoMinusChunks(aData, bData)...)
deletedChunks = append(deletedChunks, filer.DoMinusChunks(aMeta, bMeta)...)
newChunks = append(newChunks, filer.DoMinusChunks(bData, aData)...)
newChunks = append(newChunks, filer.DoMinusChunks(bMeta, aMeta)...)
return
}
|
