fix: clean up orphaned needles on remote.cache partial download failure (#8675)

When remote.cache downloads a file in parallel chunks and a gRPC
connection drops mid-transfer, chunks already written to volume servers
were not cleaned up. Since the filer metadata was never updated, these
needles became orphaned — invisible to volume.vacuum and never
referenced by the filer. On subsequent cache cycles the file was still
treated as uncached, creating more orphans each attempt.

Call DeleteUncommittedChunks on the download-error path, matching the
cleanup already present for the metadata-update-failure path.

Fixes #8481

weed/server/filer_grpc_server_remote.go

@@ -241,6 +241,12 @@ func (fs *FilerServer) doCacheRemoteObjectToLocalCluster(ctx context.Context, re
 	})
 	chunksMu.Unlock()
 	if err != nil {
+		// Clean up any chunks that were successfully written before the error.
+		// Without this, partial downloads leave orphaned needles in volume servers
+		// that accumulate across retry cycles and cannot be reclaimed by vacuum.
+		if len(chunks) > 0 {
+			fs.filer.DeleteUncommittedChunks(ctx, chunks)
+		}
 		return nil, err
 	}