add "weed filer.cat" to read files directly from volume servers

This commit is contained in:
Chris Lu
2021-01-06 04:21:34 -08:00
parent 95ecf0c72f
commit 2b76854641
9 changed files with 147 additions and 17 deletions

View File

@@ -3,6 +3,7 @@ package filer
import (
"bytes"
"fmt"
"github.com/chrislusf/seaweedfs/weed/wdclient"
"io"
"math"
"time"
@@ -38,7 +39,7 @@ func SeparateManifestChunks(chunks []*filer_pb.FileChunk) (manifestChunks, nonMa
return
}
func ResolveChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (dataChunks, manifestChunks []*filer_pb.FileChunk, manifestResolveErr error) {
func ResolveChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (dataChunks, manifestChunks []*filer_pb.FileChunk, manifestResolveErr error) {
// TODO maybe parallel this
for _, chunk := range chunks {
if !chunk.IsChunkManifest {
@@ -63,7 +64,7 @@ func ResolveChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunks []*fil
return
}
func ResolveOneChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunk *filer_pb.FileChunk) (dataChunks []*filer_pb.FileChunk, manifestResolveErr error) {
func ResolveOneChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunk *filer_pb.FileChunk) (dataChunks []*filer_pb.FileChunk, manifestResolveErr error) {
if !chunk.IsChunkManifest {
return
}
@@ -84,7 +85,7 @@ func ResolveOneChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunk *fil
}
// TODO fetch from cache for weed mount?
func fetchChunk(lookupFileIdFn LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool) ([]byte, error) {
func fetchChunk(lookupFileIdFn wdclient.LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool) ([]byte, error) {
urlStrings, err := lookupFileIdFn(fileId)
if err != nil {
glog.Errorf("operation LookupFileId %s failed, err: %v", fileId, err)

View File

@@ -4,6 +4,7 @@ import (
"bytes"
"encoding/hex"
"fmt"
"github.com/chrislusf/seaweedfs/weed/wdclient"
"math"
"sort"
"sync"
@@ -52,7 +53,7 @@ func ETagChunks(chunks []*filer_pb.FileChunk) (etag string) {
return fmt.Sprintf("%x-%d", util.Md5(bytes.Join(md5_digests, nil)), len(chunks))
}
func CompactFileChunks(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (compacted, garbage []*filer_pb.FileChunk) {
func CompactFileChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (compacted, garbage []*filer_pb.FileChunk) {
visibles, _ := NonOverlappingVisibleIntervals(lookupFileIdFn, chunks)
@@ -71,7 +72,7 @@ func CompactFileChunks(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_
return
}
func MinusChunks(lookupFileIdFn LookupFileIdFunctionType, as, bs []*filer_pb.FileChunk) (delta []*filer_pb.FileChunk, err error) {
func MinusChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, as, bs []*filer_pb.FileChunk) (delta []*filer_pb.FileChunk, err error) {
aData, aMeta, aErr := ResolveChunkManifest(lookupFileIdFn, as)
if aErr != nil {
@@ -116,7 +117,7 @@ func (cv *ChunkView) IsFullChunk() bool {
return cv.Size == cv.ChunkSize
}
func ViewFromChunks(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk, offset int64, size int64) (views []*ChunkView) {
func ViewFromChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk, offset int64, size int64) (views []*ChunkView) {
visibles, _ := NonOverlappingVisibleIntervals(lookupFileIdFn, chunks)
@@ -222,7 +223,7 @@ func MergeIntoVisibles(visibles []VisibleInterval, chunk *filer_pb.FileChunk) (n
// NonOverlappingVisibleIntervals translates the file chunk into VisibleInterval in memory
// If the file chunk content is a chunk manifest
func NonOverlappingVisibleIntervals(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (visibles []VisibleInterval, err error) {
func NonOverlappingVisibleIntervals(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (visibles []VisibleInterval, err error) {
chunks, _, err = ResolveChunkManifest(lookupFileIdFn, chunks)

View File

@@ -18,7 +18,7 @@ import (
type ChunkReadAt struct {
masterClient *wdclient.MasterClient
chunkViews []*ChunkView
lookupFileId LookupFileIdFunctionType
lookupFileId wdclient.LookupFileIdFunctionType
readerLock sync.Mutex
fileSize int64
@@ -31,9 +31,7 @@ type ChunkReadAt struct {
var _ = io.ReaderAt(&ChunkReadAt{})
var _ = io.Closer(&ChunkReadAt{})
type LookupFileIdFunctionType func(fileId string) (targetUrls []string, err error)
func LookupFn(filerClient filer_pb.FilerClient) LookupFileIdFunctionType {
func LookupFn(filerClient filer_pb.FilerClient) wdclient.LookupFileIdFunctionType {
vidCache := make(map[string]*filer_pb.Locations)
var vicCacheLock sync.RWMutex

View File

@@ -13,16 +13,16 @@ import (
"github.com/chrislusf/seaweedfs/weed/wdclient"
)
func StreamContent(masterClient *wdclient.MasterClient, w io.Writer, chunks []*filer_pb.FileChunk, offset int64, size int64) error {
func StreamContent(masterClient wdclient.HasLookupFileIdFunction, w io.Writer, chunks []*filer_pb.FileChunk, offset int64, size int64) error {
// fmt.Printf("start to stream content for chunks: %+v\n", chunks)
chunkViews := ViewFromChunks(masterClient.LookupFileId, chunks, offset, size)
chunkViews := ViewFromChunks(masterClient.GetLookupFileIdFunction(), chunks, offset, size)
fileId2Url := make(map[string][]string)
for _, chunkView := range chunkViews {
urlStrings, err := masterClient.LookupFileId(chunkView.FileId)
urlStrings, err := masterClient.GetLookupFileIdFunction()(chunkView.FileId)
if err != nil {
glog.V(1).Infof("operation LookupFileId %s failed, err: %v", chunkView.FileId, err)
return err
@@ -86,7 +86,7 @@ type ChunkStreamReader struct {
bufferOffset int64
bufferPos int
chunkIndex int
lookupFileId LookupFileIdFunctionType
lookupFileId wdclient.LookupFileIdFunctionType
}
var _ = io.ReadSeeker(&ChunkStreamReader{})