seaweedfs/weed/replication/sink/s3sink/s3_sink.go

178 lines
4.8 KiB
Go
Raw Normal View History

2018-10-04 06:36:52 +00:00
package S3Sink
import (
"fmt"
2018-10-11 07:08:13 +00:00
"github.com/aws/aws-sdk-go/aws"
"github.com/aws/aws-sdk-go/aws/credentials"
2018-10-04 06:36:52 +00:00
"github.com/aws/aws-sdk-go/aws/session"
2018-10-11 07:08:13 +00:00
"github.com/aws/aws-sdk-go/service/s3"
"github.com/aws/aws-sdk-go/service/s3/s3iface"
"github.com/aws/aws-sdk-go/service/s3/s3manager"
"strings"
"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"
2018-10-04 06:36:52 +00:00
)
type S3Sink struct {
2021-03-01 00:19:03 +00:00
conn s3iface.S3API
region string
bucket string
dir string
endpoint string
2022-02-27 10:49:31 +00:00
acl string
2021-03-01 00:19:03 +00:00
filerSource *source.FilerSource
isIncremental bool
2018-10-04 06:36:52 +00:00
}
func init() {
sink.Sinks = append(sink.Sinks, &S3Sink{})
}
func (s3sink *S3Sink) GetName() string {
2018-10-07 00:10:15 +00:00
return "s3"
2018-10-04 06:36:52 +00:00
}
func (s3sink *S3Sink) GetSinkToDirectory() string {
return s3sink.dir
}
2021-03-01 00:19:03 +00:00
func (s3sink *S3Sink) IsIncremental() bool {
return s3sink.isIncremental
}
func (s3sink *S3Sink) Initialize(configuration util.Configuration, prefix string) error {
glog.V(0).Infof("sink.s3.region: %v", configuration.GetString(prefix+"region"))
glog.V(0).Infof("sink.s3.bucket: %v", configuration.GetString(prefix+"bucket"))
glog.V(0).Infof("sink.s3.directory: %v", configuration.GetString(prefix+"directory"))
2020-04-08 00:49:00 +00:00
glog.V(0).Infof("sink.s3.endpoint: %v", configuration.GetString(prefix+"endpoint"))
2022-02-27 10:49:31 +00:00
glog.V(0).Infof("sink.s3.acl: %v", configuration.GetString(prefix+"acl"))
2021-03-01 00:19:03 +00:00
glog.V(0).Infof("sink.s3.is_incremental: %v", configuration.GetString(prefix+"is_incremental"))
s3sink.isIncremental = configuration.GetBool(prefix + "is_incremental")
2018-10-04 06:36:52 +00:00
return s3sink.initialize(
configuration.GetString(prefix+"aws_access_key_id"),
configuration.GetString(prefix+"aws_secret_access_key"),
configuration.GetString(prefix+"region"),
configuration.GetString(prefix+"bucket"),
configuration.GetString(prefix+"directory"),
2020-04-08 00:49:00 +00:00
configuration.GetString(prefix+"endpoint"),
2022-02-27 10:49:31 +00:00
configuration.GetString(prefix+"acl"),
2018-10-04 06:36:52 +00:00
)
}
func (s3sink *S3Sink) SetSourceFiler(s *source.FilerSource) {
s3sink.filerSource = s
}
2022-02-27 10:49:31 +00:00
func (s3sink *S3Sink) initialize(awsAccessKeyId, awsSecretAccessKey, region, bucket, dir, endpoint, acl string) error {
2018-10-04 06:36:52 +00:00
s3sink.region = region
s3sink.bucket = bucket
s3sink.dir = dir
2020-04-08 00:49:00 +00:00
s3sink.endpoint = endpoint
2022-02-27 10:49:31 +00:00
s3sink.acl = acl
2018-10-04 06:36:52 +00:00
config := &aws.Config{
2021-09-01 09:45:42 +00:00
Region: aws.String(s3sink.region),
Endpoint: aws.String(s3sink.endpoint),
S3ForcePathStyle: aws.Bool(true),
S3DisableContentMD5Validation: aws.Bool(true),
2018-10-04 06:36:52 +00:00
}
2019-11-19 03:24:34 +00:00
if awsAccessKeyId != "" && awsSecretAccessKey != "" {
config.Credentials = credentials.NewStaticCredentials(awsAccessKeyId, awsSecretAccessKey, "")
2018-10-04 06:36:52 +00:00
}
sess, err := session.NewSession(config)
if err != nil {
return fmt.Errorf("create aws session: %v", err)
}
s3sink.conn = s3.New(sess)
return nil
}
2020-09-09 18:21:23 +00:00
func (s3sink *S3Sink) DeleteEntry(key string, isDirectory, deleteIncludeChunks bool, signatures []int32) error {
2018-10-04 06:36:52 +00:00
2018-11-04 19:58:59 +00:00
key = cleanKey(key)
2018-10-04 06:36:52 +00:00
if isDirectory {
2022-08-27 07:09:04 +00:00
return nil
2018-10-04 06:36:52 +00:00
}
2022-08-27 07:21:57 +00:00
input := &s3.DeleteObjectInput{
Bucket: aws.String(s3sink.bucket),
Key: aws.String(key),
}
result, err := s3sink.conn.DeleteObject(input)
if err == nil {
glog.V(2).Infof("[%s] delete %s: %v", s3sink.bucket, key, result)
} else {
glog.Errorf("[%s] delete %s: %v", s3sink.bucket, key, err)
}
return err
2018-10-04 06:36:52 +00:00
}
func (s3sink *S3Sink) CreateEntry(key string, entry *filer_pb.Entry, signatures []int32) (err error) {
2018-11-04 19:58:59 +00:00
key = cleanKey(key)
2018-10-04 08:14:44 +00:00
if entry.IsDirectory {
return nil
}
reader := filer.NewFileReader(s3sink.filerSource, entry)
fileSize := int64(filer.FileSize(entry))
partSize := int64(8 * 1024 * 1024) // The minimum/default allowed part size is 5MB
for partSize*1000 < fileSize {
partSize *= 4
2018-10-04 06:36:52 +00:00
}
// Create an uploader with the session and custom options
uploader := s3manager.NewUploaderWithClient(s3sink.conn, func(u *s3manager.Uploader) {
u.PartSize = partSize
u.Concurrency = 8
})
// process tagging
tags := ""
if true {
for k, v := range entry.Extended {
if len(tags) > 0 {
tags = tags + "&"
}
tags = tags + k + "=" + string(v)
}
}
// Upload the file to S3.
_, err = uploader.Upload(&s3manager.UploadInput{
Bucket: aws.String(s3sink.bucket),
Key: aws.String(key),
Body: reader,
Tagging: aws.String(tags),
})
2018-10-04 06:36:52 +00:00
return
2018-10-04 06:36:52 +00:00
}
2020-09-09 18:21:23 +00:00
func (s3sink *S3Sink) UpdateEntry(key string, oldEntry *filer_pb.Entry, newParentPath string, newEntry *filer_pb.Entry, deleteIncludeChunks bool, signatures []int32) (foundExistingEntry bool, err error) {
2018-11-04 19:58:59 +00:00
key = cleanKey(key)
return true, s3sink.CreateEntry(key, newEntry, signatures)
2018-10-04 06:36:52 +00:00
}
2018-11-04 19:58:59 +00:00
func cleanKey(key string) string {
if strings.HasPrefix(key, "/") {
key = key[1:]
}
return key
}