Repository navigation
Expand file tree
/
Copy pathfs-v1.go
More file actions
996 lines (847 loc) · 29.2 KB
/
Copy pathfs-v1.go
File metadata and controls
996 lines (847 loc) · 29.2 KB
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
/*
* Minio Cloud Storage, (C) 2016, 2017 Minio, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package cmd
import (
"crypto/md5"
"encoding/hex"
"fmt"
"hash"
"io"
"io/ioutil"
"os"
"os/signal"
"path/filepath"
"sort"
"syscall"
"github.com/minio/minio/pkg/lock"
"github.com/minio/sha256-simd"
)
// fsObjects - Implements fs object layer.
type fsObjects struct {
// Path to be exported over S3 API.
fsPath string
// Unique value to be used for all
// temporary transactions.
fsUUID string
// FS rw pool.
rwPool *fsIOPool
// ListObjects pool management.
listPool *treeWalkPool
// To manage the appendRoutine go0routines
bgAppend *backgroundAppend
}
// Initializes meta volume on all the fs path.
func initMetaVolumeFS(fsPath, fsUUID string) error {
// This happens for the first time, but keep this here since this
// is the only place where it can be made less expensive
// optimizing all other calls. Create minio meta volume,
// if it doesn't exist yet.
metaBucketPath := pathJoin(fsPath, minioMetaBucket)
if err := mkdirAll(metaBucketPath, 0777); err != nil {
return err
}
metaTmpPath := pathJoin(fsPath, minioMetaTmpBucket, fsUUID)
if err := mkdirAll(metaTmpPath, 0777); err != nil {
return err
}
metaMultipartPath := pathJoin(fsPath, minioMetaMultipartBucket)
return mkdirAll(metaMultipartPath, 0777)
}
// Migrate FS object is a place holder code for all
// FS format migrations.
func migrateFSObject(fsPath, fsUUID string) (err error) {
// Writing message here is important for servers being upgraded.
log.Println("Please do not stop the server.")
ch := make(chan os.Signal)
defer signal.Stop(ch)
defer close(ch)
signal.Notify(ch, syscall.SIGINT, syscall.SIGTERM)
go func() {
for {
_, ok := <-ch
if !ok {
break
}
log.Println("Please wait server is being upgraded..")
}
}()
return migrateFSFormatV1ToV2(fsPath, fsUUID)
}
// List all buckets at meta bucket prefix in `.minio.sys/buckets/` path.
// This is implemented to avoid a bug on windows with using readDir().
func fsReaddirMetaBuckets(fsPath string) ([]string, error) {
f, err := os.Open(preparePath(pathJoin(fsPath, minioMetaBucket, bucketConfigPrefix)))
if err != nil {
if os.IsNotExist(err) {
return nil, errFileNotFound
} else if os.IsPermission(err) {
return nil, errFileAccessDenied
}
return nil, err
}
return f.Readdirnames(-1)
}
var bucketMetadataConfigs = []string{
bucketNotificationConfig,
bucketListenerConfig,
bucketPolicyConfig,
}
// Attempts to migrate old object metadata files to newer format
//
// i.e
// -------------------------------------------------------
// .minio.sys/buckets/<bucket_name>/<object_path>/fs.json - V1
// -------------------------------------------------------
// .minio.sys/buckets/<bucket_name>/objects/<object_path>/fs.json - V2
// -------------------------------------------------------
//
func migrateFSFormatV1ToV2(fsPath, fsUUID string) (err error) {
metaBucket := pathJoin(fsPath, minioMetaBucket, bucketConfigPrefix)
var buckets []string
buckets, err = fsReaddirMetaBuckets(fsPath)
if err != nil && err != errFileNotFound {
return err
}
// Migrate all buckets present.
for _, bucket := range buckets {
// Temporary bucket of form .UUID-bucket.
tmpBucket := fmt.Sprintf(".%s-%s", fsUUID, bucket)
// Rename existing bucket as `.UUID-bucket`.
if err = fsRenameFile(pathJoin(metaBucket, bucket), pathJoin(metaBucket, tmpBucket)); err != nil {
return err
}
// Create a new bucket name with name as `bucket`.
if err = fsMkdir(pathJoin(metaBucket, bucket)); err != nil {
return err
}
/// Rename all bucket metadata files to newly created `bucket`.
for _, bucketMetaFile := range bucketMetadataConfigs {
if err = fsRenameFile(pathJoin(metaBucket, tmpBucket, bucketMetaFile),
pathJoin(metaBucket, bucket, bucketMetaFile)); err != nil {
if errorCause(err) != errFileNotFound {
return err
}
}
}
// Finally rename the temporary bucket to `bucket/objects` directory.
if err = fsRenameFile(pathJoin(metaBucket, tmpBucket),
pathJoin(metaBucket, bucket, objectMetaPrefix)); err != nil {
if errorCause(err) != errFileNotFound {
return err
}
}
}
log.Printf("Migrating bucket metadata format from \"%s\" to newer format \"%s\"... completed successfully.", fsFormatV1, fsFormatV2)
// If all goes well we return success.
return nil
}
// newFSObjectLayer - initialize new fs object layer.
func newFSObjectLayer(fsPath string) (ObjectLayer, error) {
if fsPath == "" {
return nil, errInvalidArgument
}
// Disallow relative paths, figure out absolute paths.
fsPath, err := filepath.Abs(fsPath)
if err != nil {
return nil, err
}
fi, err := osStat(preparePath(fsPath))
if err == nil {
if !fi.IsDir() {
return nil, syscall.ENOTDIR
}
}
if os.IsNotExist(err) {
// Disk not found create it.
err = mkdirAll(fsPath, 0777)
if err != nil {
return nil, err
}
}
// Assign a new UUID for FS minio mode. Each server instance
// gets its own UUID for temporary file transaction.
fsUUID := mustGetUUID()
// Initialize meta volume, if volume already exists ignores it.
if err = initMetaVolumeFS(fsPath, fsUUID); err != nil {
return nil, fmt.Errorf("Unable to initialize '.minio.sys' meta volume, %s", err)
}
// Initialize fs objects.
fs := &fsObjects{
fsPath: fsPath,
fsUUID: fsUUID,
rwPool: &fsIOPool{
readersMap: make(map[string]*lock.RLockedFile),
},
listPool: newTreeWalkPool(globalLookupTimeout),
bgAppend: &backgroundAppend{
infoMap: make(map[string]bgAppendPartsInfo),
},
}
// Initialize `format.json`.
if err = initFormatFS(fsPath, fsUUID); err != nil {
return nil, err
}
// Once initialized hold read lock for the entire operation
// of filesystem backend.
if _, err = fs.rwPool.Open(pathJoin(fsPath, minioMetaBucket, fsFormatJSONFile)); err != nil {
return nil, err
}
// Initialize and load bucket policies.
err = initBucketPolicies(fs)
if err != nil {
return nil, fmt.Errorf("Unable to load all bucket policies. %s", err)
}
// Initialize a new event notifier.
err = initEventNotifier(fs)
if err != nil {
return nil, fmt.Errorf("Unable to initialize event notification. %s", err)
}
// Return successfully initialized object layer.
return fs, nil
}
// Should be called when process shuts down.
func (fs fsObjects) Shutdown() error {
// Close the format.json read lock.
fs.rwPool.Close(pathJoin(fs.fsPath, minioMetaBucket, fsFormatJSONFile))
// Cleanup and delete tmp uuid.
return fsRemoveAll(pathJoin(fs.fsPath, minioMetaTmpBucket, fs.fsUUID))
}
// StorageInfo - returns underlying storage statistics.
func (fs fsObjects) StorageInfo() StorageInfo {
info, err := getDiskInfo(preparePath(fs.fsPath))
errorIf(err, "Unable to get disk info %#v", fs.fsPath)
storageInfo := StorageInfo{
Total: info.Total,
Free: info.Free,
}
storageInfo.Backend.Type = FS
return storageInfo
}
/// Bucket operations
// getBucketDir - will convert incoming bucket names to
// corresponding valid bucket names on the backend in a platform
// compatible way for all operating systems.
func (fs fsObjects) getBucketDir(bucket string) (string, error) {
// Verify if bucket is valid.
if !IsValidBucketName(bucket) {
return "", traceError(BucketNameInvalid{Bucket: bucket})
}
bucketDir := pathJoin(fs.fsPath, bucket)
return bucketDir, nil
}
func (fs fsObjects) statBucketDir(bucket string) (os.FileInfo, error) {
bucketDir, err := fs.getBucketDir(bucket)
if err != nil {
return nil, err
}
st, err := fsStatDir(bucketDir)
if err != nil {
return nil, err
}
return st, nil
}
// MakeBucket - create a new bucket, returns if it
// already exists.
func (fs fsObjects) MakeBucket(bucket string) error {
bucketDir, err := fs.getBucketDir(bucket)
if err != nil {
return toObjectErr(err, bucket)
}
if err = fsMkdir(bucketDir); err != nil {
return toObjectErr(err, bucket)
}
return nil
}
// GetBucketInfo - fetch bucket metadata info.
func (fs fsObjects) GetBucketInfo(bucket string) (BucketInfo, error) {
st, err := fs.statBucketDir(bucket)
if err != nil {
return BucketInfo{}, toObjectErr(err, bucket)
}
// As osStat() doesn't carry other than ModTime(), use ModTime() as CreatedTime.
createdTime := st.ModTime()
return BucketInfo{
Name: bucket,
Created: createdTime,
}, nil
}
// ListBuckets - list all s3 compatible buckets (directories) at fsPath.
func (fs fsObjects) ListBuckets() ([]BucketInfo, error) {
if err := checkPathLength(fs.fsPath); err != nil {
return nil, traceError(err)
}
var bucketInfos []BucketInfo
entries, err := readDir(fs.fsPath)
if err != nil {
return nil, toObjectErr(traceError(errDiskNotFound))
}
for _, entry := range entries {
// Ignore all reserved bucket names and invalid bucket names.
if isReservedOrInvalidBucket(entry) {
continue
}
var fi os.FileInfo
fi, err = fsStatDir(pathJoin(fs.fsPath, entry))
if err != nil {
// If the directory does not exist, skip the entry.
if errorCause(err) == errVolumeNotFound {
continue
} else if errorCause(err) == errVolumeAccessDenied {
// Skip the entry if its a file.
continue
}
return nil, err
}
bucketInfos = append(bucketInfos, BucketInfo{
Name: fi.Name(),
// As osStat() doesnt carry CreatedTime, use ModTime() as CreatedTime.
Created: fi.ModTime(),
})
}
// Sort bucket infos by bucket name.
sort.Sort(byBucketName(bucketInfos))
// Succes.
return bucketInfos, nil
}
// DeleteBucket - delete a bucket and all the metadata associated
// with the bucket including pending multipart, object metadata.
func (fs fsObjects) DeleteBucket(bucket string) error {
bucketDir, err := fs.getBucketDir(bucket)
if err != nil {
return toObjectErr(err, bucket)
}
// Attempt to delete regular bucket.
if err = fsRemoveDir(bucketDir); err != nil {
return toObjectErr(err, bucket)
}
// Cleanup all the previously incomplete multiparts.
minioMetaMultipartBucketDir := pathJoin(fs.fsPath, minioMetaMultipartBucket, bucket)
if err = fsRemoveAll(minioMetaMultipartBucketDir); err != nil {
return toObjectErr(err, bucket)
}
// Cleanup all the bucket metadata.
minioMetadataBucketDir := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix, bucket)
if err = fsRemoveAll(minioMetadataBucketDir); err != nil {
return toObjectErr(err, bucket)
}
return nil
}
/// Object Operations
// CopyObject - copy object source object to destination object.
// if source object and destination object are same we only
// update metadata.
func (fs fsObjects) CopyObject(srcBucket, srcObject, dstBucket, dstObject string, metadata map[string]string) (ObjectInfo, error) {
if _, err := fs.statBucketDir(srcBucket); err != nil {
return ObjectInfo{}, toObjectErr(err, srcBucket)
}
// Stat the file to get file size.
fi, err := fsStatFile(pathJoin(fs.fsPath, srcBucket, srcObject))
if err != nil {
return ObjectInfo{}, toObjectErr(err, srcBucket, srcObject)
}
// Check if this request is only metadata update.
cpMetadataOnly := isStringEqual(pathJoin(srcBucket, srcObject), pathJoin(dstBucket, dstObject))
if cpMetadataOnly {
fsMetaPath := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix, srcBucket, objectMetaPrefix, srcObject, fsMetaJSONFile)
var wlk *lock.LockedFile
wlk, err = fs.rwPool.Write(fsMetaPath)
if err != nil {
return ObjectInfo{}, toObjectErr(traceError(err), srcBucket, srcObject)
}
// This close will allow for locks to be synchronized on `fs.json`.
defer wlk.Close()
// Save objects' metadata in `fs.json`.
fsMeta := newFSMetaV1()
fsMeta.Meta = metadata
if _, err = fsMeta.WriteTo(wlk); err != nil {
return ObjectInfo{}, toObjectErr(err, srcBucket, srcObject)
}
// Return the new object info.
return fsMeta.ToObjectInfo(srcBucket, srcObject, fi), nil
}
// Length of the file to read.
length := fi.Size()
// Initialize pipe.
pipeReader, pipeWriter := io.Pipe()
go func() {
var startOffset int64 // Read the whole file.
if gerr := fs.GetObject(srcBucket, srcObject, startOffset, length, pipeWriter); gerr != nil {
errorIf(gerr, "Unable to read %s/%s.", srcBucket, srcObject)
pipeWriter.CloseWithError(gerr)
return
}
pipeWriter.Close() // Close writer explicitly signalling we wrote all data.
}()
objInfo, err := fs.PutObject(dstBucket, dstObject, length, pipeReader, metadata, "")
if err != nil {
return ObjectInfo{}, toObjectErr(err, dstBucket, dstObject)
}
// Explicitly close the reader.
pipeReader.Close()
return objInfo, nil
}
// GetObject - reads an object from the disk.
// Supports additional parameters like offset and length
// which are synonymous with HTTP Range requests.
//
// startOffset indicates the starting read location of the object.
// length indicates the total length of the object.
func (fs fsObjects) GetObject(bucket, object string, offset int64, length int64, writer io.Writer) (err error) {
if err = checkGetObjArgs(bucket, object); err != nil {
return err
}
if _, err = fs.statBucketDir(bucket); err != nil {
return toObjectErr(err, bucket)
}
// Offset cannot be negative.
if offset < 0 {
return toObjectErr(traceError(errUnexpected), bucket, object)
}
// Writer cannot be nil.
if writer == nil {
return toObjectErr(traceError(errUnexpected), bucket, object)
}
if bucket != minioMetaBucket {
fsMetaPath := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix, bucket, objectMetaPrefix, object, fsMetaJSONFile)
_, err = fs.rwPool.Open(fsMetaPath)
if err != nil && err != errFileNotFound {
return toObjectErr(traceError(err), bucket, object)
}
defer fs.rwPool.Close(fsMetaPath)
}
// Read the object, doesn't exist returns an s3 compatible error.
fsObjPath := pathJoin(fs.fsPath, bucket, object)
reader, size, err := fsOpenFile(fsObjPath, offset)
if err != nil {
return toObjectErr(err, bucket, object)
}
defer reader.Close()
bufSize := int64(readSizeV1)
if length > 0 && bufSize > length {
bufSize = length
}
// For negative length we read everything.
if length < 0 {
length = size - offset
}
// Reply back invalid range if the input offset and length fall out of range.
if offset > size || offset+length > size {
return traceError(InvalidRange{offset, length, size})
}
// Allocate a staging buffer.
buf := make([]byte, int(bufSize))
_, err = io.CopyBuffer(writer, io.LimitReader(reader, length), buf)
return toObjectErr(traceError(err), bucket, object)
}
// getObjectInfo - wrapper for reading object metadata and constructs ObjectInfo.
func (fs fsObjects) getObjectInfo(bucket, object string) (ObjectInfo, error) {
fsMeta := fsMetaV1{}
fsMetaPath := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix, bucket, objectMetaPrefix, object, fsMetaJSONFile)
// Read `fs.json` to perhaps contend with
// parallel Put() operations.
rlk, err := fs.rwPool.Open(fsMetaPath)
if err == nil {
// Read from fs metadata only if it exists.
defer fs.rwPool.Close(fsMetaPath)
if _, rerr := fsMeta.ReadFrom(rlk.LockedFile); rerr != nil {
// `fs.json` can be empty due to previously failed
// PutObject() transaction, if we arrive at such
// a situation we just ignore and continue.
if errorCause(rerr) != io.EOF {
return ObjectInfo{}, toObjectErr(rerr, bucket, object)
}
}
}
// Ignore if `fs.json` is not available, this is true for pre-existing data.
if err != nil && err != errFileNotFound {
return ObjectInfo{}, toObjectErr(traceError(err), bucket, object)
}
// Stat the file to get file size.
fi, err := fsStatFile(pathJoin(fs.fsPath, bucket, object))
if err != nil {
return ObjectInfo{}, toObjectErr(err, bucket, object)
}
return fsMeta.ToObjectInfo(bucket, object, fi), nil
}
// GetObjectInfo - reads object metadata and replies back ObjectInfo.
func (fs fsObjects) GetObjectInfo(bucket, object string) (ObjectInfo, error) {
// This is a special case with object whose name ends with
// a slash separator, we always return object not found here.
if hasSuffix(object, slashSeparator) {
return ObjectInfo{}, toObjectErr(traceError(errFileNotFound), bucket, object)
}
if err := checkGetObjArgs(bucket, object); err != nil {
return ObjectInfo{}, err
}
if _, err := fs.statBucketDir(bucket); err != nil {
return ObjectInfo{}, toObjectErr(err, bucket)
}
return fs.getObjectInfo(bucket, object)
}
// PutObject - creates an object upon reading from the input stream
// until EOF, writes data directly to configured filesystem path.
// Additionally writes `fs.json` which carries the necessary metadata
// for future object operations.
func (fs fsObjects) PutObject(bucket string, object string, size int64, data io.Reader, metadata map[string]string, sha256sum string) (objInfo ObjectInfo, retErr error) {
var err error
// This is a special case with size as '0' and object ends with
// a slash separator, we treat it like a valid operation and
// return success.
if isObjectDir(object, size) {
return dirObjectInfo(bucket, object, size, metadata), nil
}
if err = checkPutObjectArgs(bucket, object, fs); err != nil {
return ObjectInfo{}, err
}
if _, err = fs.statBucketDir(bucket); err != nil {
return ObjectInfo{}, toObjectErr(err, bucket)
}
// No metadata is set, allocate a new one.
if metadata == nil {
metadata = make(map[string]string)
}
fsMeta := newFSMetaV1()
fsMeta.Meta = metadata
var wlk *lock.LockedFile
if bucket != minioMetaBucket {
bucketMetaDir := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix)
fsMetaPath := pathJoin(bucketMetaDir, bucket, objectMetaPrefix, object, fsMetaJSONFile)
wlk, err = fs.rwPool.Create(fsMetaPath)
if err != nil {
return ObjectInfo{}, toObjectErr(traceError(err), bucket, object)
}
// This close will allow for locks to be synchronized on `fs.json`.
defer wlk.Close()
defer func() {
// Remove meta file when PutObject encounters any error
if retErr != nil {
tmpDir := pathJoin(fs.fsPath, minioMetaTmpBucket, fs.fsUUID)
fsRemoveMeta(bucketMetaDir, fsMetaPath, tmpDir)
}
}()
}
// Uploaded object will first be written to the temporary location which will eventually
// be renamed to the actual location. It is first written to the temporary location
// so that cleaning it up will be easy if the server goes down.
tempObj := mustGetUUID()
// Initialize md5 writer.
md5Writer := md5.New()
hashWriters := []io.Writer{md5Writer}
var sha256Writer hash.Hash
if sha256sum != "" {
sha256Writer = sha256.New()
hashWriters = append(hashWriters, sha256Writer)
}
multiWriter := io.MultiWriter(hashWriters...)
// Limit the reader to its provided size if specified.
var limitDataReader io.Reader
if size > 0 {
// This is done so that we can avoid erroneous clients sending more data than the set content size.
limitDataReader = io.LimitReader(data, size)
} else {
// else we read till EOF.
limitDataReader = data
}
// Allocate a buffer to Read() from request body
bufSize := int64(readSizeV1)
if size > 0 && bufSize > size {
bufSize = size
}
buf := make([]byte, int(bufSize))
teeReader := io.TeeReader(limitDataReader, multiWriter)
fsTmpObjPath := pathJoin(fs.fsPath, minioMetaTmpBucket, fs.fsUUID, tempObj)
bytesWritten, err := fsCreateFile(fsTmpObjPath, teeReader, buf, size)
if err != nil {
fsRemoveFile(fsTmpObjPath)
errorIf(err, "Failed to create object %s/%s", bucket, object)
return ObjectInfo{}, toObjectErr(err, bucket, object)
}
// Should return IncompleteBody{} error when reader has fewer
// bytes than specified in request header.
if bytesWritten < size {
fsRemoveFile(fsTmpObjPath)
return ObjectInfo{}, traceError(IncompleteBody{})
}
// Delete the temporary object in the case of a
// failure. If PutObject succeeds, then there would be
// nothing to delete.
defer fsRemoveFile(fsTmpObjPath)
newMD5Hex := hex.EncodeToString(md5Writer.Sum(nil))
// Update the md5sum if not set with the newly calculated one.
if len(metadata["md5Sum"]) == 0 {
metadata["md5Sum"] = newMD5Hex
}
// md5Hex representation.
md5Hex := metadata["md5Sum"]
if md5Hex != "" {
if newMD5Hex != md5Hex {
// Returns md5 mismatch.
return ObjectInfo{}, traceError(BadDigest{md5Hex, newMD5Hex})
}
}
if sha256sum != "" {
newSHA256sum := hex.EncodeToString(sha256Writer.Sum(nil))
if newSHA256sum != sha256sum {
return ObjectInfo{}, traceError(SHA256Mismatch{})
}
}
// Entire object was written to the temp location, now it's safe to rename it to the actual location.
fsNSObjPath := pathJoin(fs.fsPath, bucket, object)
if err = fsRenameFile(fsTmpObjPath, fsNSObjPath); err != nil {
return ObjectInfo{}, toObjectErr(err, bucket, object)
}
if bucket != minioMetaBucket {
// Write FS metadata after a successful namespace operation.
if _, err = fsMeta.WriteTo(wlk); err != nil {
return ObjectInfo{}, toObjectErr(err, bucket, object)
}
}
// Stat the file to fetch timestamp, size.
fi, err := fsStatFile(pathJoin(fs.fsPath, bucket, object))
if err != nil {
return ObjectInfo{}, toObjectErr(err, bucket, object)
}
// Success.
return fsMeta.ToObjectInfo(bucket, object, fi), nil
}
// DeleteObject - deletes an object from a bucket, this operation is destructive
// and there are no rollbacks supported.
func (fs fsObjects) DeleteObject(bucket, object string) error {
if err := checkDelObjArgs(bucket, object); err != nil {
return err
}
if _, err := fs.statBucketDir(bucket); err != nil {
return toObjectErr(err, bucket)
}
minioMetaBucketDir := pathJoin(fs.fsPath, minioMetaBucket)
fsMetaPath := pathJoin(minioMetaBucketDir, bucketMetaPrefix, bucket, objectMetaPrefix, object, fsMetaJSONFile)
if bucket != minioMetaBucket {
rwlk, lerr := fs.rwPool.Write(fsMetaPath)
if lerr == nil {
// This close will allow for fs locks to be synchronized on `fs.json`.
defer rwlk.Close()
}
if lerr != nil && lerr != errFileNotFound {
return toObjectErr(traceError(lerr), bucket, object)
}
}
// Delete the object.
if err := fsDeleteFile(pathJoin(fs.fsPath, bucket), pathJoin(fs.fsPath, bucket, object)); err != nil {
return toObjectErr(err, bucket, object)
}
if bucket != minioMetaBucket {
// Delete the metadata object.
err := fsDeleteFile(minioMetaBucketDir, fsMetaPath)
if err != nil && errorCause(err) != errFileNotFound {
return toObjectErr(err, bucket, object)
}
}
return nil
}
// list of all errors that can be ignored in tree walk operation in FS
var fsTreeWalkIgnoredErrs = append(baseIgnoredErrs, []error{
errFileNotFound,
errVolumeNotFound,
}...)
// Returns function "listDir" of the type listDirFunc.
// isLeaf - is used by listDir function to check if an entry
// is a leaf or non-leaf entry.
func (fs fsObjects) listDirFactory(isLeaf isLeafFunc) listDirFunc {
// listDir - lists all the entries at a given prefix and given entry in the prefix.
listDir := func(bucket, prefixDir, prefixEntry string) (entries []string, delayIsLeaf bool, err error) {
entries, err = readDir(pathJoin(fs.fsPath, bucket, prefixDir))
if err != nil {
return nil, false, err
}
entries, delayIsLeaf = filterListEntries(bucket, prefixDir, entries, prefixEntry, isLeaf)
return entries, delayIsLeaf, nil
}
// Return list factory instance.
return listDir
}
// getObjectETag is a helper function, which returns only the md5sum
// of the file on the disk.
func (fs fsObjects) getObjectETag(bucket, entry string) (string, error) {
fsMetaPath := pathJoin(fs.fsPath, minioMetaBucket, bucketMetaPrefix, bucket, objectMetaPrefix, entry, fsMetaJSONFile)
// Read `fs.json` to perhaps contend with
// parallel Put() operations.
rlk, err := fs.rwPool.Open(fsMetaPath)
// Ignore if `fs.json` is not available, this is true for pre-existing data.
if err != nil && err != errFileNotFound {
return "", toObjectErr(traceError(err), bucket, entry)
}
// If file is not found, we don't need to proceed forward.
if err == errFileNotFound {
return "", nil
}
// Read from fs metadata only if it exists.
defer fs.rwPool.Close(fsMetaPath)
fsMetaBuf, err := ioutil.ReadAll(rlk.LockedFile)
if err != nil {
// `fs.json` can be empty due to previously failed
// PutObject() transaction, if we arrive at such
// a situation we just ignore and continue.
if errorCause(err) != io.EOF {
return "", toObjectErr(err, bucket, entry)
}
}
fsMetaMap := parseFSMetaMap(fsMetaBuf)
return fsMetaMap["md5Sum"], nil
}
// ListObjects - list all objects at prefix upto maxKeys., optionally delimited by '/'. Maintains the list pool
// state for future re-entrant list requests.
func (fs fsObjects) ListObjects(bucket, prefix, marker, delimiter string, maxKeys int) (ListObjectsInfo, error) {
if err := checkListObjsArgs(bucket, prefix, marker, delimiter, fs); err != nil {
return ListObjectsInfo{}, err
}
if _, err := fs.statBucketDir(bucket); err != nil {
return ListObjectsInfo{}, err
}
// With max keys of zero we have reached eof, return right here.
if maxKeys == 0 {
return ListObjectsInfo{}, nil
}
// For delimiter and prefix as '/' we do not list anything at all
// since according to s3 spec we stop at the 'delimiter'
// along // with the prefix. On a flat namespace with 'prefix'
// as '/' we don't have any entries, since all the keys are
// of form 'keyName/...'
if delimiter == slashSeparator && prefix == slashSeparator {
return ListObjectsInfo{}, nil
}
// Over flowing count - reset to maxObjectList.
if maxKeys < 0 || maxKeys > maxObjectList {
maxKeys = maxObjectList
}
// Default is recursive, if delimiter is set then list non recursive.
recursive := true
if delimiter == slashSeparator {
recursive = false
}
// Convert entry to ObjectInfo
entryToObjectInfo := func(entry string) (objInfo ObjectInfo, err error) {
if hasSuffix(entry, slashSeparator) {
// Object name needs to be full path.
objInfo.Name = entry
objInfo.IsDir = true
return
}
// Protect reading `fs.json`.
objectLock := globalNSMutex.NewNSLock(bucket, entry)
objectLock.RLock()
var md5Sum string
md5Sum, err = fs.getObjectETag(bucket, entry)
objectLock.RUnlock()
if err != nil {
return ObjectInfo{}, err
}
// Stat the file to get file size.
var fi os.FileInfo
fi, err = fsStatFile(pathJoin(fs.fsPath, bucket, entry))
if err != nil {
return ObjectInfo{}, toObjectErr(err, bucket, entry)
}
// Success.
return ObjectInfo{
Name: entry,
Bucket: bucket,
Size: fi.Size(),
ModTime: fi.ModTime(),
IsDir: fi.IsDir(),
MD5Sum: md5Sum,
}, nil
}
heal := false // true only for xl.ListObjectsHeal()
walkResultCh, endWalkCh := fs.listPool.Release(listParams{bucket, recursive, marker, prefix, heal})
if walkResultCh == nil {
endWalkCh = make(chan struct{})
isLeaf := func(bucket, object string) bool {
// bucket argument is unused as we don't need to StatFile
// to figure if it's a file, just need to check that the
// object string does not end with "/".
return !hasSuffix(object, slashSeparator)
}
listDir := fs.listDirFactory(isLeaf)
walkResultCh = startTreeWalk(bucket, prefix, marker, recursive, listDir, isLeaf, endWalkCh)
}
var objInfos []ObjectInfo
var eof bool
var nextMarker string
// List until maxKeys requested.
for i := 0; i < maxKeys; {
walkResult, ok := <-walkResultCh
if !ok {
// Closed channel.
eof = true
break
}
// For any walk error return right away.
if walkResult.err != nil {
// File not found is a valid case.
if errorCause(walkResult.err) == errFileNotFound {
return ListObjectsInfo{}, nil
}
return ListObjectsInfo{}, toObjectErr(walkResult.err, bucket, prefix)
}
objInfo, err := entryToObjectInfo(walkResult.entry)
if err != nil {
return ListObjectsInfo{}, nil
}
nextMarker = objInfo.Name
objInfos = append(objInfos, objInfo)
if walkResult.end {
eof = true
break
}
i++
}
// Save list routine for the next marker if we haven't reached EOF.
params := listParams{bucket, recursive, nextMarker, prefix, heal}
if !eof {
fs.listPool.Set(params, walkResultCh, endWalkCh)
}
result := ListObjectsInfo{IsTruncated: !eof}
for _, objInfo := range objInfos {
result.NextMarker = objInfo.Name
if objInfo.IsDir {
result.Prefixes = append(result.Prefixes, objInfo.Name)
continue
}
result.Objects = append(result.Objects, objInfo)
}
// Success.
return result, nil
}
// HealObject - no-op for fs. Valid only for XL.
func (fs fsObjects) HealObject(bucket, object string) (int, int, error) {
return 0, 0, traceError(NotImplemented{})
}
// HealBucket - no-op for fs, Valid only for XL.
func (fs fsObjects) HealBucket(bucket string) error {
return traceError(NotImplemented{})
}
// ListObjectsHeal - list all objects to be healed. Valid only for XL
func (fs fsObjects) ListObjectsHeal(bucket, prefix, marker, delimiter string, maxKeys int) (ListObjectsInfo, error) {
return ListObjectsInfo{}, traceError(NotImplemented{})
}
// ListBucketsHeal - list all buckets to be healed. Valid only for XL
func (fs fsObjects) ListBucketsHeal() ([]BucketInfo, error) {
return []BucketInfo{}, traceError(NotImplemented{})
}
func (fs fsObjects) ListUploadsHeal(bucket, prefix, marker, uploadIDMarker,
delimiter string, maxUploads int) (ListMultipartsInfo, error) {
return ListMultipartsInfo{}, traceError(NotImplemented{})
}