mirror of
https://github.com/minio/minio.git
synced 2024-12-24 06:05:55 -05:00
avoid concurrent reads and writes to opts.UserDefined (#14862)
do not modify opts.UserDefined after object-handler has set all the necessary values, any mutation needed should be done on a copy of this value not directly. As there are other pieces of code that access opts.UserDefined concurrently this becomes challenging. fixes #14856
This commit is contained in:
parent
2b7e75e079
commit
c7df1ffc6f
@ -884,6 +884,7 @@ func (c *diskCache) put(ctx context.Context, bucket, object string, data io.Read
|
||||
}
|
||||
|
||||
if err := os.MkdirAll(cachePath, 0o777); err != nil {
|
||||
removeAll(cachePath)
|
||||
return oi, err
|
||||
}
|
||||
metadata := cloneMSS(opts.UserDefined)
|
||||
@ -892,6 +893,7 @@ func (c *diskCache) put(ctx context.Context, bucket, object string, data io.Read
|
||||
if globalCacheKMS != nil {
|
||||
reader, err = newCacheEncryptReader(data, bucket, object, metadata)
|
||||
if err != nil {
|
||||
removeAll(cachePath)
|
||||
return oi, err
|
||||
}
|
||||
actualSize, _ = sio.EncryptedSize(uint64(size))
|
||||
|
@ -33,7 +33,6 @@ import (
|
||||
"github.com/minio/minio/internal/config/cache"
|
||||
"github.com/minio/minio/internal/disk"
|
||||
"github.com/minio/minio/internal/hash"
|
||||
xhttp "github.com/minio/minio/internal/http"
|
||||
"github.com/minio/minio/internal/logger"
|
||||
"github.com/minio/minio/internal/sync/errgroup"
|
||||
"github.com/minio/pkg/wildcard"
|
||||
@ -739,35 +738,30 @@ func (c *cacheObjects) PutObject(ctx context.Context, bucket, object string, r *
|
||||
// Initialize pipe to stream data to cache
|
||||
rPipe, wPipe := io.Pipe()
|
||||
infoCh := make(chan ObjectInfo)
|
||||
errorCh := make(chan error)
|
||||
go func() {
|
||||
defer close(infoCh)
|
||||
info, err := putObjectFn(ctx, bucket, object, NewPutObjReader(hashReader), opts)
|
||||
if err != nil {
|
||||
close(infoCh)
|
||||
pipeReader.CloseWithError(err)
|
||||
rPipe.CloseWithError(err)
|
||||
errorCh <- err
|
||||
return
|
||||
pipeReader.CloseWithError(err)
|
||||
rPipe.CloseWithError(err)
|
||||
if err == nil {
|
||||
infoCh <- info
|
||||
}
|
||||
close(errorCh)
|
||||
infoCh <- info
|
||||
}()
|
||||
|
||||
go func() {
|
||||
_, err := dcache.put(lkctx.Context(), bucket, object, rPipe, r.Size(), nil, opts, false, false)
|
||||
if err != nil {
|
||||
rPipe.CloseWithError(err)
|
||||
return
|
||||
logger.LogIf(lkctx.Context(), err)
|
||||
}
|
||||
// We do not care about errors to cached backend.
|
||||
rPipe.Close()
|
||||
}()
|
||||
|
||||
mwriter := cacheMultiWriter(pipeWriter, wPipe)
|
||||
_, err = io.Copy(mwriter, r)
|
||||
pipeWriter.Close()
|
||||
wPipe.Close()
|
||||
|
||||
if err != nil {
|
||||
err = <-errorCh
|
||||
return ObjectInfo{}, err
|
||||
}
|
||||
info := <-infoCh
|
||||
@ -803,8 +797,7 @@ func (c *cacheObjects) uploadObject(ctx context.Context, oi ObjectInfo) {
|
||||
return
|
||||
}
|
||||
var opts ObjectOptions
|
||||
opts.UserDefined = make(map[string]string)
|
||||
opts.UserDefined[xhttp.ContentMD5] = oi.UserDefined["content-md5"]
|
||||
opts.UserDefined = cloneMSS(oi.UserDefined)
|
||||
objInfo, err := c.InnerPutObjectFn(ctx, oi.Bucket, oi.Name, NewPutObjReader(hashReader), opts)
|
||||
wbCommitStatus := CommitComplete
|
||||
size := objInfo.Size
|
||||
|
@ -286,8 +286,10 @@ func (er erasureObjects) ListMultipartUploads(ctx context.Context, bucket, objec
|
||||
// disks. `uploads.json` carries metadata regarding on-going multipart
|
||||
// operation(s) on the object.
|
||||
func (er erasureObjects) newMultipartUpload(ctx context.Context, bucket string, object string, opts ObjectOptions) (string, error) {
|
||||
userDefined := cloneMSS(opts.UserDefined)
|
||||
|
||||
onlineDisks := er.getDisks()
|
||||
parityDrives := globalStorageClass.GetParityForSC(opts.UserDefined[xhttp.AmzStorageClass])
|
||||
parityDrives := globalStorageClass.GetParityForSC(userDefined[xhttp.AmzStorageClass])
|
||||
if parityDrives <= 0 {
|
||||
parityDrives = er.defaultParityCount
|
||||
}
|
||||
@ -308,7 +310,7 @@ func (er erasureObjects) newMultipartUpload(ctx context.Context, bucket string,
|
||||
}
|
||||
}
|
||||
if parityOrig != parityDrives {
|
||||
opts.UserDefined[minIOErasureUpgraded] = strconv.Itoa(parityOrig) + "->" + strconv.Itoa(parityDrives)
|
||||
userDefined[minIOErasureUpgraded] = strconv.Itoa(parityOrig) + "->" + strconv.Itoa(parityDrives)
|
||||
}
|
||||
|
||||
dataDrives := len(onlineDisks) - parityDrives
|
||||
@ -336,8 +338,8 @@ func (er erasureObjects) newMultipartUpload(ctx context.Context, bucket string,
|
||||
}
|
||||
|
||||
// Guess content-type from the extension if possible.
|
||||
if opts.UserDefined["content-type"] == "" {
|
||||
opts.UserDefined["content-type"] = mimedb.TypeByExtension(path.Ext(object))
|
||||
if userDefined["content-type"] == "" {
|
||||
userDefined["content-type"] = mimedb.TypeByExtension(path.Ext(object))
|
||||
}
|
||||
|
||||
modTime := opts.MTime
|
||||
@ -352,7 +354,7 @@ func (er erasureObjects) newMultipartUpload(ctx context.Context, bucket string,
|
||||
for index := range partsMetadata {
|
||||
partsMetadata[index].Fresh = true
|
||||
partsMetadata[index].ModTime = modTime
|
||||
partsMetadata[index].Metadata = opts.UserDefined
|
||||
partsMetadata[index].Metadata = userDefined
|
||||
}
|
||||
|
||||
uploadID := mustGetUUID()
|
||||
@ -375,10 +377,6 @@ func (er erasureObjects) newMultipartUpload(ctx context.Context, bucket string,
|
||||
func (er erasureObjects) NewMultipartUpload(ctx context.Context, bucket, object string, opts ObjectOptions) (string, error) {
|
||||
auditObjectErasureSet(ctx, object, &er)
|
||||
|
||||
// No metadata is set, allocate a new one.
|
||||
if opts.UserDefined == nil {
|
||||
opts.UserDefined = make(map[string]string)
|
||||
}
|
||||
return er.newMultipartUpload(ctx, bucket, object, opts)
|
||||
}
|
||||
|
||||
|
@ -842,17 +842,14 @@ func (er erasureObjects) putObject(ctx context.Context, bucket string, object st
|
||||
|
||||
data := r.Reader
|
||||
|
||||
// No metadata is set, allocate a new one.
|
||||
if opts.UserDefined == nil {
|
||||
opts.UserDefined = make(map[string]string)
|
||||
}
|
||||
userDefined := cloneMSS(opts.UserDefined)
|
||||
|
||||
storageDisks := er.getDisks()
|
||||
|
||||
parityDrives := len(storageDisks) / 2
|
||||
if !opts.MaxParity {
|
||||
// Get parity and data drive count based on storage class metadata
|
||||
parityDrives = globalStorageClass.GetParityForSC(opts.UserDefined[xhttp.AmzStorageClass])
|
||||
parityDrives = globalStorageClass.GetParityForSC(userDefined[xhttp.AmzStorageClass])
|
||||
if parityDrives <= 0 {
|
||||
parityDrives = er.defaultParityCount
|
||||
}
|
||||
@ -890,7 +887,7 @@ func (er erasureObjects) putObject(ctx context.Context, bucket string, object st
|
||||
parityDrives = len(storageDisks) / 2
|
||||
}
|
||||
if parityOrig != parityDrives {
|
||||
opts.UserDefined[minIOErasureUpgraded] = strconv.Itoa(parityOrig) + "->" + strconv.Itoa(parityDrives)
|
||||
userDefined[minIOErasureUpgraded] = strconv.Itoa(parityOrig) + "->" + strconv.Itoa(parityDrives)
|
||||
}
|
||||
}
|
||||
dataDrives := len(storageDisks) - parityDrives
|
||||
@ -1066,13 +1063,13 @@ func (er erasureObjects) putObject(ctx context.Context, bucket string, object st
|
||||
Hash: bitrotWriterSum(w),
|
||||
})
|
||||
}
|
||||
if opts.UserDefined["etag"] == "" {
|
||||
opts.UserDefined["etag"] = r.MD5CurrentHexString()
|
||||
if userDefined["etag"] == "" {
|
||||
userDefined["etag"] = r.MD5CurrentHexString()
|
||||
}
|
||||
|
||||
// Guess content-type from the extension if possible.
|
||||
if opts.UserDefined["content-type"] == "" {
|
||||
opts.UserDefined["content-type"] = mimedb.TypeByExtension(path.Ext(object))
|
||||
if userDefined["content-type"] == "" {
|
||||
userDefined["content-type"] = mimedb.TypeByExtension(path.Ext(object))
|
||||
}
|
||||
|
||||
modTime := opts.MTime
|
||||
@ -1083,7 +1080,7 @@ func (er erasureObjects) putObject(ctx context.Context, bucket string, object st
|
||||
// Fill all the necessary metadata.
|
||||
// Update `xl.meta` content on each disks.
|
||||
for index := range partsMetadata {
|
||||
partsMetadata[index].Metadata = opts.UserDefined
|
||||
partsMetadata[index].Metadata = userDefined
|
||||
partsMetadata[index].Size = n
|
||||
partsMetadata[index].ModTime = modTime
|
||||
}
|
||||
|
@ -364,6 +364,7 @@ func (l *s3EncObjects) GetObjectInfo(ctx context.Context, bucket string, object
|
||||
// CopyObject copies an object from source bucket to a destination bucket.
|
||||
func (l *s3EncObjects) CopyObject(ctx context.Context, srcBucket string, srcObject string, dstBucket string, dstObject string, srcInfo minio.ObjectInfo, s, d minio.ObjectOptions) (objInfo minio.ObjectInfo, err error) {
|
||||
cpSrcDstSame := path.Join(srcBucket, srcObject) == path.Join(dstBucket, dstObject)
|
||||
userDefined := minio.CloneMSS(srcInfo.UserDefined)
|
||||
if cpSrcDstSame {
|
||||
var gwMeta gwMetaV1
|
||||
if s.ServerSideEncryption != nil && d.ServerSideEncryption != nil &&
|
||||
@ -378,16 +379,16 @@ func (l *s3EncObjects) CopyObject(ctx context.Context, srcBucket string, srcObje
|
||||
d.ServerSideEncryption.Marshal(header)
|
||||
}
|
||||
for k, v := range header {
|
||||
srcInfo.UserDefined[k] = v[0]
|
||||
userDefined[k] = v[0]
|
||||
}
|
||||
gwMeta.Meta = srcInfo.UserDefined
|
||||
gwMeta.Meta = userDefined
|
||||
if err = l.writeGWMetadata(ctx, dstBucket, getDareMetaPath(dstObject), gwMeta, minio.ObjectOptions{}); err != nil {
|
||||
return objInfo, minio.ErrorRespToObjectError(err)
|
||||
}
|
||||
return gwMeta.ToObjectInfo(dstBucket, dstObject), nil
|
||||
}
|
||||
}
|
||||
dstOpts := minio.ObjectOptions{ServerSideEncryption: d.ServerSideEncryption, UserDefined: srcInfo.UserDefined}
|
||||
dstOpts := minio.ObjectOptions{ServerSideEncryption: d.ServerSideEncryption, UserDefined: userDefined}
|
||||
return l.PutObject(ctx, dstBucket, dstObject, srcInfo.PutObjReader, dstOpts)
|
||||
}
|
||||
|
||||
|
@ -502,17 +502,20 @@ func (l *s3Objects) GetObjectInfo(ctx context.Context, bucket string, object str
|
||||
// PutObject creates a new object with the incoming data,
|
||||
func (l *s3Objects) PutObject(ctx context.Context, bucket string, object string, r *minio.PutObjReader, opts minio.ObjectOptions) (objInfo minio.ObjectInfo, err error) {
|
||||
data := r.Reader
|
||||
|
||||
userDefined := minio.CloneMSS(opts.UserDefined)
|
||||
|
||||
var tagMap map[string]string
|
||||
if tagstr, ok := opts.UserDefined[xhttp.AmzObjectTagging]; ok && tagstr != "" {
|
||||
if tagstr, ok := userDefined[xhttp.AmzObjectTagging]; ok && tagstr != "" {
|
||||
tagObj, err := tags.ParseObjectTags(tagstr)
|
||||
if err != nil {
|
||||
return objInfo, minio.ErrorRespToObjectError(err, bucket, object)
|
||||
}
|
||||
tagMap = tagObj.ToMap()
|
||||
delete(opts.UserDefined, xhttp.AmzObjectTagging)
|
||||
delete(userDefined, xhttp.AmzObjectTagging)
|
||||
}
|
||||
putOpts := miniogo.PutObjectOptions{
|
||||
UserMetadata: opts.UserDefined,
|
||||
UserMetadata: userDefined,
|
||||
ServerSideEncryption: opts.ServerSideEncryption,
|
||||
UserTags: tagMap,
|
||||
// Content-Md5 is needed for buckets with object locking,
|
||||
@ -529,7 +532,7 @@ func (l *s3Objects) PutObject(ctx context.Context, bucket string, object string,
|
||||
ETag: ui.ETag,
|
||||
Size: ui.Size,
|
||||
Key: object,
|
||||
Metadata: minio.ToMinioClientObjectInfoMetadata(opts.UserDefined),
|
||||
Metadata: minio.ToMinioClientObjectInfoMetadata(userDefined),
|
||||
}
|
||||
|
||||
return minio.FromMinioClientObjectInfo(bucket, oi), nil
|
||||
@ -544,8 +547,9 @@ func (l *s3Objects) CopyObject(ctx context.Context, srcBucket string, srcObject
|
||||
// metadata input is already a trickled down value from interpreting x-amz-metadata-directive at
|
||||
// handler layer. So what we have right now is supposed to be applied on the destination object anyways.
|
||||
// So preserve it by adding "REPLACE" directive to save all the metadata set by CopyObject API.
|
||||
srcInfo.UserDefined["x-amz-metadata-directive"] = "REPLACE"
|
||||
srcInfo.UserDefined["x-amz-copy-source-if-match"] = srcInfo.ETag
|
||||
userDefined := minio.CloneMSS(srcInfo.UserDefined)
|
||||
userDefined["x-amz-metadata-directive"] = "REPLACE"
|
||||
userDefined["x-amz-copy-source-if-match"] = srcInfo.ETag
|
||||
header := make(http.Header)
|
||||
if srcOpts.ServerSideEncryption != nil {
|
||||
encrypt.SSECopy(srcOpts.ServerSideEncryption).Marshal(header)
|
||||
@ -556,10 +560,10 @@ func (l *s3Objects) CopyObject(ctx context.Context, srcBucket string, srcObject
|
||||
}
|
||||
|
||||
for k, v := range header {
|
||||
srcInfo.UserDefined[k] = v[0]
|
||||
userDefined[k] = v[0]
|
||||
}
|
||||
|
||||
if _, err = l.Client.CopyObject(ctx, srcBucket, srcObject, dstBucket, dstObject, srcInfo.UserDefined, miniogo.CopySrcOptions{}, miniogo.PutObjectOptions{}); err != nil {
|
||||
if _, err = l.Client.CopyObject(ctx, srcBucket, srcObject, dstBucket, dstObject, userDefined, miniogo.CopySrcOptions{}, miniogo.PutObjectOptions{}); err != nil {
|
||||
return objInfo, minio.ErrorRespToObjectError(err, srcBucket, srcObject)
|
||||
}
|
||||
return l.GetObjectInfo(ctx, dstBucket, dstObject, dstOpts)
|
||||
@ -605,17 +609,18 @@ func (l *s3Objects) ListMultipartUploads(ctx context.Context, bucket string, pre
|
||||
// NewMultipartUpload upload object in multiple parts
|
||||
func (l *s3Objects) NewMultipartUpload(ctx context.Context, bucket string, object string, o minio.ObjectOptions) (uploadID string, err error) {
|
||||
var tagMap map[string]string
|
||||
if tagStr, ok := o.UserDefined[xhttp.AmzObjectTagging]; ok {
|
||||
userDefined := minio.CloneMSS(o.UserDefined)
|
||||
if tagStr, ok := userDefined[xhttp.AmzObjectTagging]; ok {
|
||||
tagObj, err := tags.Parse(tagStr, true)
|
||||
if err != nil {
|
||||
return uploadID, minio.ErrorRespToObjectError(err, bucket, object)
|
||||
}
|
||||
tagMap = tagObj.ToMap()
|
||||
delete(o.UserDefined, xhttp.AmzObjectTagging)
|
||||
delete(userDefined, xhttp.AmzObjectTagging)
|
||||
}
|
||||
// Create PutObject options
|
||||
opts := miniogo.PutObjectOptions{
|
||||
UserMetadata: o.UserDefined,
|
||||
UserMetadata: userDefined,
|
||||
ServerSideEncryption: o.ServerSideEncryption,
|
||||
UserTags: tagMap,
|
||||
}
|
||||
@ -645,9 +650,8 @@ func (l *s3Objects) CopyObjectPart(ctx context.Context, srcBucket, srcObject, de
|
||||
if srcOpts.CheckPrecondFn != nil && srcOpts.CheckPrecondFn(srcInfo) {
|
||||
return minio.PartInfo{}, minio.PreConditionFailed{}
|
||||
}
|
||||
srcInfo.UserDefined = map[string]string{
|
||||
"x-amz-copy-source-if-match": srcInfo.ETag,
|
||||
}
|
||||
userDefined := minio.CloneMSS(srcInfo.UserDefined)
|
||||
userDefined["x-amz-copy-source-if-match"] = srcInfo.ETag
|
||||
header := make(http.Header)
|
||||
if srcOpts.ServerSideEncryption != nil {
|
||||
encrypt.SSECopy(srcOpts.ServerSideEncryption).Marshal(header)
|
||||
@ -657,11 +661,11 @@ func (l *s3Objects) CopyObjectPart(ctx context.Context, srcBucket, srcObject, de
|
||||
dstOpts.ServerSideEncryption.Marshal(header)
|
||||
}
|
||||
for k, v := range header {
|
||||
srcInfo.UserDefined[k] = v[0]
|
||||
userDefined[k] = v[0]
|
||||
}
|
||||
|
||||
completePart, err := l.Client.CopyObjectPart(ctx, srcBucket, srcObject, destBucket, destObject,
|
||||
uploadID, partID, startOffset, length, srcInfo.UserDefined)
|
||||
uploadID, partID, startOffset, length, userDefined)
|
||||
if err != nil {
|
||||
return p, minio.ErrorRespToObjectError(err, srcBucket, srcObject)
|
||||
}
|
||||
|
@ -130,6 +130,9 @@ func getWriteQuorum(drive int) int {
|
||||
return quorum
|
||||
}
|
||||
|
||||
// CloneMSS is an exposed function of cloneMSS for gateway usage.
|
||||
var CloneMSS = cloneMSS
|
||||
|
||||
// cloneMSS will clone a map[string]string.
|
||||
// If input is nil an empty map is returned, not nil.
|
||||
func cloneMSS(v map[string]string) map[string]string {
|
||||
|
Loading…
Reference in New Issue
Block a user