rclone/backend/drive/upload.go
Nick Craig-Wood 6757244918 drive: use multipart resumable uploads for streaming and uploads in mount
Before this change we used non multipart uploads for files of unknown
size (streaming and uploads in mount).  This is slower and less
reliable and is not recommended by Google for files smaller than 5MB.

After this change we use multipart resumable uploads for all files of
unknown length.  This will use an extra transaction so is less
efficient for files under the chunk size, however the natural
buffering in the operations.Rcat call specified by
`--streaming-upload-cutoff` will overcome this.

See: https://forum.rclone.org/t/upload-behaviour-and-speed-when-using-vfs-cache/9920/
2020-01-17 22:03:10 +00:00

244 lines
7.6 KiB
Go

// Upload for drive
//
// Docs
// Resumable upload: https://developers.google.com/drive/web/manage-uploads#resumable
// Best practices: https://developers.google.com/drive/web/manage-uploads#best-practices
// Files insert: https://developers.google.com/drive/v2/reference/files/insert
// Files update: https://developers.google.com/drive/v2/reference/files/update
//
// This contains code adapted from google.golang.org/api (C) the GO AUTHORS
package drive
import (
"bytes"
"context"
"encoding/json"
"fmt"
"io"
"net/http"
"net/url"
"strconv"
"github.com/rclone/rclone/fs"
"github.com/rclone/rclone/fs/fserrors"
"github.com/rclone/rclone/lib/readers"
"google.golang.org/api/drive/v3"
"google.golang.org/api/googleapi"
)
const (
// statusResumeIncomplete is the code returned by the Google uploader when the transfer is not yet complete.
statusResumeIncomplete = 308
)
// resumableUpload is used by the generated APIs to provide resumable uploads.
// It is not used by developers directly.
type resumableUpload struct {
f *Fs
remote string
// URI is the resumable resource destination provided by the server after specifying "&uploadType=resumable".
URI string
// Media is the object being uploaded.
Media io.Reader
// MediaType defines the media type, e.g. "image/jpeg".
MediaType string
// ContentLength is the full size of the object being uploaded.
ContentLength int64
// Return value
ret *drive.File
}
// Upload the io.Reader in of size bytes with contentType and info
func (f *Fs) Upload(ctx context.Context, in io.Reader, size int64, contentType, fileID, remote string, info *drive.File) (*drive.File, error) {
params := url.Values{
"alt": {"json"},
"uploadType": {"resumable"},
"fields": {partialFields},
}
params.Set("supportsAllDrives", "true")
if f.opt.KeepRevisionForever {
params.Set("keepRevisionForever", "true")
}
urls := "https://www.googleapis.com/upload/drive/v3/files"
method := "POST"
if fileID != "" {
params.Set("setModifiedDate", "true")
urls += "/{fileId}"
method = "PATCH"
}
urls += "?" + params.Encode()
var res *http.Response
var err error
err = f.pacer.Call(func() (bool, error) {
var body io.Reader
body, err = googleapi.WithoutDataWrapper.JSONReader(info)
if err != nil {
return false, err
}
var req *http.Request
req, err = http.NewRequest(method, urls, body)
if err != nil {
return false, err
}
req = req.WithContext(ctx) // go1.13 can use NewRequestWithContext
googleapi.Expand(req.URL, map[string]string{
"fileId": fileID,
})
req.Header.Set("Content-Type", "application/json; charset=UTF-8")
req.Header.Set("X-Upload-Content-Type", contentType)
if size >= 0 {
req.Header.Set("X-Upload-Content-Length", fmt.Sprintf("%v", size))
}
res, err = f.client.Do(req)
if err == nil {
defer googleapi.CloseBody(res)
err = googleapi.CheckResponse(res)
}
return f.shouldRetry(err)
})
if err != nil {
return nil, err
}
loc := res.Header.Get("Location")
rx := &resumableUpload{
f: f,
remote: remote,
URI: loc,
Media: in,
MediaType: contentType,
ContentLength: size,
}
return rx.Upload(ctx)
}
// Make an http.Request for the range passed in
func (rx *resumableUpload) makeRequest(ctx context.Context, start int64, body io.ReadSeeker, reqSize int64) *http.Request {
req, _ := http.NewRequest("POST", rx.URI, body)
req = req.WithContext(ctx) // go1.13 can use NewRequestWithContext
req.ContentLength = reqSize
totalSize := "*"
if rx.ContentLength >= 0 {
totalSize = strconv.FormatInt(rx.ContentLength, 10)
}
if reqSize != 0 {
req.Header.Set("Content-Range", fmt.Sprintf("bytes %v-%v/%v", start, start+reqSize-1, totalSize))
} else {
req.Header.Set("Content-Range", fmt.Sprintf("bytes */%v", totalSize))
}
req.Header.Set("Content-Type", rx.MediaType)
return req
}
// Transfer a chunk - caller must call googleapi.CloseBody(res) if err == nil || res != nil
func (rx *resumableUpload) transferChunk(ctx context.Context, start int64, chunk io.ReadSeeker, chunkSize int64) (int, error) {
_, _ = chunk.Seek(0, io.SeekStart)
req := rx.makeRequest(ctx, start, chunk, chunkSize)
res, err := rx.f.client.Do(req)
if err != nil {
return 599, err
}
defer googleapi.CloseBody(res)
if res.StatusCode == statusResumeIncomplete {
return res.StatusCode, nil
}
err = googleapi.CheckResponse(res)
if err != nil {
return res.StatusCode, err
}
// When the entire file upload is complete, the server
// responds with an HTTP 201 Created along with any metadata
// associated with this resource. If this request had been
// updating an existing entity rather than creating a new one,
// the HTTP response code for a completed upload would have
// been 200 OK.
//
// So parse the response out of the body. We aren't expecting
// any other 2xx codes, so we parse it unconditionally on
// StatusCode
if err = json.NewDecoder(res.Body).Decode(&rx.ret); err != nil {
return 598, err
}
return res.StatusCode, nil
}
// Upload uploads the chunks from the input
// It retries each chunk using the pacer and --low-level-retries
func (rx *resumableUpload) Upload(ctx context.Context) (*drive.File, error) {
start := int64(0)
var StatusCode int
var err error
buf := make([]byte, int(rx.f.opt.ChunkSize))
for finished := false; !finished; {
var reqSize int64
var chunk io.ReadSeeker
if rx.ContentLength >= 0 {
// If size known use repeatable reader for smoother bwlimit
if start >= rx.ContentLength {
break
}
reqSize = rx.ContentLength - start
if reqSize >= int64(rx.f.opt.ChunkSize) {
reqSize = int64(rx.f.opt.ChunkSize)
}
chunk = readers.NewRepeatableLimitReaderBuffer(rx.Media, buf, reqSize)
} else {
// If size unknown read into buffer
var n int
n, err = readers.ReadFill(rx.Media, buf)
if err == io.EOF {
// Send the last chunk with the correct ContentLength
// otherwise Google doesn't know we've finished
rx.ContentLength = start + int64(n)
finished = true
} else if err != nil {
return nil, err
}
reqSize = int64(n)
chunk = bytes.NewReader(buf[:reqSize])
}
// Transfer the chunk
err = rx.f.pacer.Call(func() (bool, error) {
fs.Debugf(rx.remote, "Sending chunk %d length %d", start, reqSize)
StatusCode, err = rx.transferChunk(ctx, start, chunk, reqSize)
again, err := rx.f.shouldRetry(err)
if StatusCode == statusResumeIncomplete || StatusCode == http.StatusCreated || StatusCode == http.StatusOK {
again = false
err = nil
}
return again, err
})
if err != nil {
return nil, err
}
start += reqSize
}
// Resume or retry uploads that fail due to connection interruptions or
// any 5xx errors, including:
//
// 500 Internal Server Error
// 502 Bad Gateway
// 503 Service Unavailable
// 504 Gateway Timeout
//
// Use an exponential backoff strategy if any 5xx server error is
// returned when resuming or retrying upload requests. These errors can
// occur if a server is getting overloaded. Exponential backoff can help
// alleviate these kinds of problems during periods of high volume of
// requests or heavy network traffic. Other kinds of requests should not
// be handled by exponential backoff but you can still retry a number of
// them. When retrying these requests, limit the number of times you
// retry them. For example your code could limit to ten retries or less
// before reporting an error.
//
// Handle 404 Not Found errors when doing resumable uploads by starting
// the entire upload over from the beginning.
if rx.ret == nil {
return nil, fserrors.RetryErrorf("Incomplete upload - retry, last error %d", StatusCode)
}
return rx.ret, nil
}