apache/hadoop · error · UnsupportedFileSystemException

FileSystem '{}'is not a ViewFileSystem.

Error message

FileSystem '{}'is not a ViewFileSystem.

What it means

ViewFileSystemUtil.getStatus(FileSystem, Path) computes per-mount-point FsStatus but only accepts ViewFileSystem or ViewFileSystemOverloadScheme instances. Any other FileSystem (plain DistributedFileSystem, RawLocalFileSystem, S3A, HarFs, ...) fails fast with UnsupportedFileSystemException("FileSystem '<uri>'is not a ViewFileSystem.") before any status is read.

Source

Thrown at hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/viewfs/ViewFileSystemUtil.java:110

   *
   *  "/dept/eng"           Not a mount point, but a valid   (1), (2), (3), (4)
   *                         internal dir in the mount tree
   *                         and resolved down to "/" path.
   *
   *  "/erp"                Doesn't match or leads to or
   *                         over any valid mount points     None
   *
   *
   * @param fileSystem - ViewFileSystem on which mount point exists
   * @param path - URI for which FsStatus is requested
   * @return Map of ViewFsMountPoint and FsStatus
   * @throws IOException raised on errors performing I/O.
   */
  public static Map<MountPoint, FsStatus> getStatus(
      FileSystem fileSystem, Path path) throws IOException {
    if (!(isViewFileSystem(fileSystem)
        || isViewFileSystemOverloadScheme(fileSystem))) {
      throw new UnsupportedFileSystemException("FileSystem '"
          + fileSystem.getUri() + "'is not a ViewFileSystem.");
    }
    ViewFileSystem viewFileSystem = (ViewFileSystem) fileSystem;
    String viewFsUriPath = viewFileSystem.getUriPath(path);
    boolean isPathOverMountPoint = false;
    boolean isPathLeadingToMountPoint = false;
    boolean isPathIncludesAllMountPoint = false;
    Map<MountPoint, FsStatus> mountPointMap = new HashMap<>();
    for (MountPoint mountPoint : viewFileSystem.getMountPoints()) {
      String[] mountPointPathComponents = InodeTree.breakIntoPathComponents(
          mountPoint.getMountedOnPath().toString());
      String[] incomingPathComponents =
          InodeTree.breakIntoPathComponents(viewFsUriPath);

      int pathCompIndex;
      for (pathCompIndex = 0; pathCompIndex < mountPointPathComponents.length &&
          pathCompIndex < incomingPathComponents.length; pathCompIndex++) {
        if (!mountPointPathComponents[pathCompIndex].equals(

View on GitHub (pinned to 2add963021)

Solutions

1. Guard the call with ViewFileSystemUtil.isViewFileSystem(fs) || ViewFileSystemUtil.isViewFileSystemOverloadScheme(fs) and branch to fs.getStatus() otherwise
2. Construct the viewfs FileSystem explicitly: FileSystem.get(new URI("viewfs://cluster"), conf)
3. Ensure the tool is pointed at a deployment whose defaultFS is the viewfs mount table

Example fix

// before
Map<MountPoint, FsStatus> m = ViewFileSystemUtil.getStatus(fs, path); // throws for hdfs://

// after
if (ViewFileSystemUtil.isViewFileSystem(fs)
    || ViewFileSystemUtil.isViewFileSystemOverloadScheme(fs)) {
  Map<MountPoint, FsStatus> m = ViewFileSystemUtil.getStatus(fs, path);
} else {
  FsStatus s = fs.getStatus(path); // single-FS fallback path
}

if (!ViewFileSystemUtil.isViewFileSystem(fs)
    && !ViewFileSystemUtil.isViewFileSystemOverloadScheme(fs)) {
  FsStatus s = fs.getStatus(path); // single-FS path
} else {
  Map<MountPoint, FsStatus> m = ViewFileSystemUtil.getStatus(fs, path);
}

static boolean isViewFsCapable(FileSystem fs) {
  // ViewFileSystemOverloadScheme extends ViewFileSystem, so one check covers both
  return fs instanceof org.apache.hadoop.fs.viewfs.ViewFileSystem;
}

try {
  statuses = ViewFileSystemUtil.getStatus(fs, path);
} catch (UnsupportedFileSystemException e) {
  statuses = Map.of(); // fall back to fs.getStatus(path) for plain filesystems
}

When it happens

Trigger: Calling ViewFileSystemUtil.getStatus(FileSystem.get(conf), path) when the Configuration resolves to a non-viewfs scheme; helper code shared between unit tests (local FS) and production (viewfs) passing whatever FileSystem it was handed.

Common situations: Cluster-capacity UIs and ops tooling that assume the defaultFS is always viewfs:// (federation) but are run on a single-cluster gateway where defaultFS=hdfs://; test harnesses using LocalFileSystem through the same code path.

Related errors

• getStatus on path `{}' is not within a mount point
• getXAttrs on path `{}' is not within a mount point
• listXAttrs on path `{}' is not within a mount point
• getQuotaUsage on path `{}' is not within a mount point
• getStoragePolicy on path `{}' is not within a mount point

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/87f98bbcc03375c6. Report an issue: GitHub.