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
|
/*
* Minio Go Library for Amazon S3 Compatible Cloud Storage (C) 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 minio
import (
"fmt"
"io"
"net/http"
"sort"
"strings"
"github.com/minio/minio-go/pkg/s3utils"
)
// PutObjectStreaming using AWS streaming signature V4
func (c Client) PutObjectStreaming(bucketName, objectName string, reader io.Reader) (n int64, err error) {
return c.PutObjectWithProgress(bucketName, objectName, reader, nil, nil)
}
// putObjectMultipartStream - upload a large object using
// multipart upload and streaming signature for signing payload.
// Comprehensive put object operation involving multipart uploads.
//
// Following code handles these types of readers.
//
// - *os.File
// - *minio.Object
// - Any reader which has a method 'ReadAt()'
//
func (c Client) putObjectMultipartStream(bucketName, objectName string,
reader io.Reader, size int64, metadata map[string][]string, progress io.Reader) (n int64, err error) {
// Verify if reader is *minio.Object, *os.File or io.ReaderAt.
// NOTE: Verification of object is kept for a specific purpose
// while it is going to be duck typed similar to io.ReaderAt.
// It is to indicate that *minio.Object implements io.ReaderAt.
// and such a functionality is used in the subsequent code path.
if isFile(reader) || !isObject(reader) && isReadAt(reader) {
n, err = c.putObjectMultipartStreamFromReadAt(bucketName, objectName, reader.(io.ReaderAt), size, metadata, progress)
} else {
n, err = c.putObjectMultipartStreamNoChecksum(bucketName, objectName, reader, size, metadata, progress)
}
if err != nil {
errResp := ToErrorResponse(err)
// Verify if multipart functionality is not available, if not
// fall back to single PutObject operation.
if errResp.Code == "AccessDenied" && strings.Contains(errResp.Message, "Access Denied") {
// Verify if size of reader is greater than '5GiB'.
if size > maxSinglePutObjectSize {
return 0, ErrEntityTooLarge(size, maxSinglePutObjectSize, bucketName, objectName)
}
// Fall back to uploading as single PutObject operation.
return c.putObjectNoChecksum(bucketName, objectName, reader, size, metadata, progress)
}
}
return n, err
}
// uploadedPartRes - the response received from a part upload.
type uploadedPartRes struct {
Error error // Any error encountered while uploading the part.
PartNum int // Number of the part uploaded.
Size int64 // Size of the part uploaded.
Part *ObjectPart
}
type uploadPartReq struct {
PartNum int // Number of the part uploaded.
Part *ObjectPart // Size of the part uploaded.
}
// putObjectMultipartFromReadAt - Uploads files bigger than 64MiB.
// Supports all readers which implements io.ReaderAt interface
// (ReadAt method).
//
// NOTE: This function is meant to be used for all readers which
// implement io.ReaderAt which allows us for resuming multipart
// uploads but reading at an offset, which would avoid re-read the
// data which was already uploaded. Internally this function uses
// temporary files for staging all the data, these temporary files are
// cleaned automatically when the caller i.e http client closes the
// stream after uploading all the contents successfully.
func (c Client) putObjectMultipartStreamFromReadAt(bucketName, objectName string,
reader io.ReaderAt, size int64, metadata map[string][]string, progress io.Reader) (n int64, err error) {
// Input validation.
if err = s3utils.CheckValidBucketName(bucketName); err != nil {
return 0, err
}
if err = s3utils.CheckValidObjectName(objectName); err != nil {
return 0, err
}
// Calculate the optimal parts info for a given size.
totalPartsCount, partSize, lastPartSize, err := optimalPartInfo(size)
if err != nil {
return 0, err
}
// Initiate a new multipart upload.
uploadID, err := c.newUploadID(bucketName, objectName, metadata)
if err != nil {
return 0, err
}
// Aborts the multipart upload in progress, if the
// function returns any error, since we do not resume
// we should purge the parts which have been uploaded
// to relinquish storage space.
defer func() {
if err != nil {
c.abortMultipartUpload(bucketName, objectName, uploadID)
}
}()
// Total data read and written to server. should be equal to 'size' at the end of the call.
var totalUploadedSize int64
// Complete multipart upload.
var complMultipartUpload completeMultipartUpload
// Declare a channel that sends the next part number to be uploaded.
// Buffered to 10000 because thats the maximum number of parts allowed
// by S3.
uploadPartsCh := make(chan uploadPartReq, 10000)
// Declare a channel that sends back the response of a part upload.
// Buffered to 10000 because thats the maximum number of parts allowed
// by S3.
uploadedPartsCh := make(chan uploadedPartRes, 10000)
// Used for readability, lastPartNumber is always totalPartsCount.
lastPartNumber := totalPartsCount
// Send each part number to the channel to be processed.
for p := 1; p <= totalPartsCount; p++ {
uploadPartsCh <- uploadPartReq{PartNum: p, Part: nil}
}
close(uploadPartsCh)
// Receive each part number from the channel allowing three parallel uploads.
for w := 1; w <= totalWorkers; w++ {
go func(partSize int64) {
// Each worker will draw from the part channel and upload in parallel.
for uploadReq := range uploadPartsCh {
// If partNumber was not uploaded we calculate the missing
// part offset and size. For all other part numbers we
// calculate offset based on multiples of partSize.
readOffset := int64(uploadReq.PartNum-1) * partSize
// As a special case if partNumber is lastPartNumber, we
// calculate the offset based on the last part size.
if uploadReq.PartNum == lastPartNumber {
readOffset = (size - lastPartSize)
partSize = lastPartSize
}
// Get a section reader on a particular offset.
sectionReader := newHook(io.NewSectionReader(reader, readOffset, partSize), progress)
// Proceed to upload the part.
var objPart ObjectPart
objPart, err = c.uploadPart(bucketName, objectName, uploadID,
sectionReader, uploadReq.PartNum,
nil, nil, partSize, metadata)
if err != nil {
uploadedPartsCh <- uploadedPartRes{
Size: 0,
Error: err,
}
// Exit the goroutine.
return
}
// Save successfully uploaded part metadata.
uploadReq.Part = &objPart
// Send successful part info through the channel.
uploadedPartsCh <- uploadedPartRes{
Size: objPart.Size,
PartNum: uploadReq.PartNum,
Part: uploadReq.Part,
Error: nil,
}
}
}(partSize)
}
// Gather the responses as they occur and update any
// progress bar.
for u := 1; u <= totalPartsCount; u++ {
uploadRes := <-uploadedPartsCh
if uploadRes.Error != nil {
return totalUploadedSize, uploadRes.Error
}
// Retrieve each uploaded part and store it to be completed.
// part, ok := partsInfo[uploadRes.PartNum]
part := uploadRes.Part
if part == nil {
return 0, ErrInvalidArgument(fmt.Sprintf("Missing part number %d", uploadRes.PartNum))
}
// Update the totalUploadedSize.
totalUploadedSize += uploadRes.Size
// Store the parts to be completed in order.
complMultipartUpload.Parts = append(complMultipartUpload.Parts, CompletePart{
ETag: part.ETag,
PartNumber: part.PartNumber,
})
}
// Verify if we uploaded all the data.
if totalUploadedSize != size {
return totalUploadedSize, ErrUnexpectedEOF(totalUploadedSize, size, bucketName, objectName)
}
// Sort all completed parts.
sort.Sort(completedParts(complMultipartUpload.Parts))
_, err = c.completeMultipartUpload(bucketName, objectName, uploadID, complMultipartUpload)
if err != nil {
return totalUploadedSize, err
}
// Return final size.
return totalUploadedSize, nil
}
func (c Client) putObjectMultipartStreamNoChecksum(bucketName, objectName string,
reader io.Reader, size int64, metadata map[string][]string, progress io.Reader) (n int64, err error) {
// Input validation.
if err = s3utils.CheckValidBucketName(bucketName); err != nil {
return 0, err
}
if err = s3utils.CheckValidObjectName(objectName); err != nil {
return 0, err
}
// Calculate the optimal parts info for a given size.
totalPartsCount, partSize, lastPartSize, err := optimalPartInfo(size)
if err != nil {
return 0, err
}
// Initiates a new multipart request
uploadID, err := c.newUploadID(bucketName, objectName, metadata)
if err != nil {
return 0, err
}
// Aborts the multipart upload if the function returns
// any error, since we do not resume we should purge
// the parts which have been uploaded to relinquish
// storage space.
defer func() {
if err != nil {
c.abortMultipartUpload(bucketName, objectName, uploadID)
}
}()
// Total data read and written to server. should be equal to 'size' at the end of the call.
var totalUploadedSize int64
// Initialize parts uploaded map.
partsInfo := make(map[int]ObjectPart)
// Part number always starts with '1'.
var partNumber int
for partNumber = 1; partNumber <= totalPartsCount; partNumber++ {
// Update progress reader appropriately to the latest offset
// as we read from the source.
hookReader := newHook(reader, progress)
// Proceed to upload the part.
if partNumber == totalPartsCount {
partSize = lastPartSize
}
var objPart ObjectPart
objPart, err = c.uploadPart(bucketName, objectName, uploadID,
io.LimitReader(hookReader, partSize),
partNumber, nil, nil, partSize, metadata)
if err != nil {
return totalUploadedSize, err
}
// Save successfully uploaded part metadata.
partsInfo[partNumber] = objPart
// Save successfully uploaded size.
totalUploadedSize += partSize
}
// Verify if we uploaded all the data.
if size > 0 {
if totalUploadedSize != size {
return totalUploadedSize, ErrUnexpectedEOF(totalUploadedSize, size, bucketName, objectName)
}
}
// Complete multipart upload.
var complMultipartUpload completeMultipartUpload
// Loop over total uploaded parts to save them in
// Parts array before completing the multipart request.
for i := 1; i < partNumber; i++ {
part, ok := partsInfo[i]
if !ok {
return 0, ErrInvalidArgument(fmt.Sprintf("Missing part number %d", i))
}
complMultipartUpload.Parts = append(complMultipartUpload.Parts, CompletePart{
ETag: part.ETag,
PartNumber: part.PartNumber,
})
}
// Sort all completed parts.
sort.Sort(completedParts(complMultipartUpload.Parts))
_, err = c.completeMultipartUpload(bucketName, objectName, uploadID, complMultipartUpload)
if err != nil {
return totalUploadedSize, err
}
// Return final size.
return totalUploadedSize, nil
}
// putObjectNoChecksum special function used Google Cloud Storage. This special function
// is used for Google Cloud Storage since Google's multipart API is not S3 compatible.
func (c Client) putObjectNoChecksum(bucketName, objectName string, reader io.Reader, size int64, metaData map[string][]string, progress io.Reader) (n int64, err error) {
// Input validation.
if err := s3utils.CheckValidBucketName(bucketName); err != nil {
return 0, err
}
if err := s3utils.CheckValidObjectName(objectName); err != nil {
return 0, err
}
// Size -1 is only supported on Google Cloud Storage, we error
// out in all other situations.
if size < 0 && !s3utils.IsGoogleEndpoint(c.endpointURL) {
return 0, ErrEntityTooSmall(size, bucketName, objectName)
}
if size > 0 {
if isReadAt(reader) && !isObject(reader) {
reader = io.NewSectionReader(reader.(io.ReaderAt), 0, size)
}
}
// Update progress reader appropriately to the latest offset as we
// read from the source.
readSeeker := newHook(reader, progress)
// This function does not calculate sha256 and md5sum for payload.
// Execute put object.
st, err := c.putObjectDo(bucketName, objectName, readSeeker, nil, nil, size, metaData)
if err != nil {
return 0, err
}
if st.Size != size {
return 0, ErrUnexpectedEOF(st.Size, size, bucketName, objectName)
}
return size, nil
}
// putObjectDo - executes the put object http operation.
// NOTE: You must have WRITE permissions on a bucket to add an object to it.
func (c Client) putObjectDo(bucketName, objectName string, reader io.Reader, md5Sum []byte, sha256Sum []byte, size int64, metaData map[string][]string) (ObjectInfo, error) {
// Input validation.
if err := s3utils.CheckValidBucketName(bucketName); err != nil {
return ObjectInfo{}, err
}
if err := s3utils.CheckValidObjectName(objectName); err != nil {
return ObjectInfo{}, err
}
// Set headers.
customHeader := make(http.Header)
// Set metadata to headers
for k, v := range metaData {
if len(v) > 0 {
customHeader.Set(k, v[0])
}
}
// If Content-Type is not provided, set the default application/octet-stream one
if v, ok := metaData["Content-Type"]; !ok || len(v) == 0 {
customHeader.Set("Content-Type", "application/octet-stream")
}
// Populate request metadata.
reqMetadata := requestMetadata{
bucketName: bucketName,
objectName: objectName,
customHeader: customHeader,
contentBody: reader,
contentLength: size,
contentMD5Bytes: md5Sum,
contentSHA256Bytes: sha256Sum,
}
// Execute PUT an objectName.
resp, err := c.executeMethod("PUT", reqMetadata)
defer closeResponse(resp)
if err != nil {
return ObjectInfo{}, err
}
if resp != nil {
if resp.StatusCode != http.StatusOK {
return ObjectInfo{}, httpRespToErrorResponse(resp, bucketName, objectName)
}
}
var objInfo ObjectInfo
// Trim off the odd double quotes from ETag in the beginning and end.
objInfo.ETag = strings.TrimPrefix(resp.Header.Get("ETag"), "\"")
objInfo.ETag = strings.TrimSuffix(objInfo.ETag, "\"")
// A success here means data was written to server successfully.
objInfo.Size = size
// Return here.
return objInfo, nil
}
|