filer: let NewChunkStreamReaderFromFiler fall back to the master on a cache miss (#11964)

refreshUrls invalidates every cached location of a volume after a failed
read and re-looks it up through the reader's lookup function. With the
cache-only masterClient.LookupFileId, one failed read to one replica left
the volume missing from the filer's cache until restart, so metadata-log
replay skipped it. Use GetLookupFileIdFunction like the other read paths.

Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: Chris Lu <chrislusf@users.noreply.github.com>

weed/filer/stream.go

@@ -415,11 +415,7 @@ func doNewChunkStreamReader(ctx context.Context, lookupFileIdFn wdclient.LookupF
 
 func NewChunkStreamReaderFromFiler(ctx context.Context, masterClient *wdclient.MasterClient, chunks []*filer_pb.FileChunk) *ChunkStreamReader {
 
-	lookupFileIdFn := func(ctx context.Context, fileId string) (targetUrl []string, err error) {
-		return masterClient.LookupFileId(ctx, fileId)
-	}
-
-	return doNewChunkStreamReader(ctx, lookupFileIdFn, chunks)
+	return doNewChunkStreamReader(ctx, masterClient.GetLookupFileIdFunction(), chunks)
 }
 
 // NewChunkStreamReaderFromLookup creates a ChunkStreamReader from a lookup function.

weed/filer/stream_from_filer_lookup_test.go

@@ -0,0 +1,101 @@
+package filer
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+	"time"
+
+	"google.golang.org/grpc"
+	"google.golang.org/grpc/credentials/insecure"
+
+	"github.com/seaweedfs/seaweedfs/weed/cluster"
+	"github.com/seaweedfs/seaweedfs/weed/pb"
+	"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
+	"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
+	"github.com/seaweedfs/seaweedfs/weed/wdclient"
+)
+
+// lookupOnlyMaster announces no volumes on KeepConnected, so the client's
+// cache stays empty, but answers LookupVolume with the one volume server.
+type lookupOnlyMaster struct {
+	master_pb.UnimplementedSeaweedServer
+	volumeUrl string
+}
+
+func (m *lookupOnlyMaster) KeepConnected(stream master_pb.Seaweed_KeepConnectedServer) error {
+	if _, err := stream.Recv(); err != nil {
+		return err
+	}
+	if err := stream.Send(&master_pb.KeepConnectedResponse{}); err != nil {
+		return err
+	}
+	<-stream.Context().Done()
+	return nil
+}
+
+func (m *lookupOnlyMaster) LookupVolume(ctx context.Context, req *master_pb.LookupVolumeRequest) (*master_pb.LookupVolumeResponse, error) {
+	resp := &master_pb.LookupVolumeResponse{}
+	for _, id := range req.VolumeOrFileIds {
+		resp.VolumeIdLocations = append(resp.VolumeIdLocations, &master_pb.LookupVolumeResponse_VolumeIdLocation{
+			VolumeOrFileId: id,
+			Locations:      []*master_pb.Location{{Url: m.volumeUrl, PublicUrl: m.volumeUrl}},
+		})
+	}
+	return resp, nil
+}
+
+// A filer's cached volume map can lose a volume that still exists: a failed
+// read invalidates every cached location of it, and the master only announces
+// changes. Readers built from the filer's MasterClient must then ask the master
+// instead of failing until the process restarts.
+func TestChunkStreamReaderFromFilerAsksMasterOnCacheMiss(t *testing.T) {
+	content := []byte("metadata log chunk")
+	volume := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		_, _ = w.Write(content)
+	}))
+	t.Cleanup(volume.Close)
+
+	master := &lookupOnlyMaster{volumeUrl: strings.TrimPrefix(volume.URL, "http://")}
+	lis, err := net.Listen("tcp", "127.0.0.1:0")
+	if err != nil {
+		t.Fatalf("listen: %v", err)
+	}
+	grpcServer := grpc.NewServer()
+	master_pb.RegisterSeaweedServer(grpcServer, master)
+	go func() { _ = grpcServer.Serve(lis) }()
+	t.Cleanup(grpcServer.Stop)
+	_, port, err := net.SplitHostPort(lis.Addr().String())
+	if err != nil {
+		t.Fatalf("split listener address: %v", err)
+	}
+
+	mc := wdclient.NewMasterClient(
+		grpc.WithTransportCredentials(insecure.NewCredentials()),
+		"", cluster.FilerType, pb.ServerAddress("localhost:0"), "", "",
+		*pb.NewServiceDiscoveryFromMap(map[string]pb.ServerAddress{"m": pb.ServerAddress(fmt.Sprintf("127.0.0.1:0.%s", port))}),
+	)
+	connecting, stopConnecting := context.WithCancel(context.Background())
+	t.Cleanup(stopConnecting)
+	go mc.KeepConnectedToMaster(connecting)
+	waiting, cancelWaiting := context.WithTimeout(context.Background(), 20*time.Second)
+	defer cancelWaiting()
+	mc.WaitUntilConnected(waiting)
+	if waiting.Err() != nil {
+		t.Fatal("the master client never connected to the fake master")
+	}
+
+	chunks := []*filer_pb.FileChunk{{FileId: "1,0123456789", Offset: 0, Size: uint64(len(content))}}
+	got, err := io.ReadAll(NewChunkStreamReaderFromFiler(context.Background(), mc, chunks))
+	if err != nil {
+		t.Fatalf("read with volume 1 missing from the cache: %v", err)
+	}
+	if string(got) != string(content) {
+		t.Fatalf("read %q, want %q", got, content)
+	}
+}