apache/hadoop · error · NotInMountpointException

listXAttrs on path `{}' is not within a mount point

Error message

listXAttrs on path `{}' is not within a mount point

What it means

listXAttrs on ViewFileSystem's InternalDirOfViewFs throws NotInMountpointException when invoked on an internal mount-table directory — a virtual node (like "/" or an intermediate container dir) that has no child FileSystem behind it. The exception is unchecked (extends UnsupportedOperationException), so it escapes typical IOException-only handling.

Source

Thrown at hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/viewfs/ViewFileSystem.java:1856

    @Override
    public byte[] getXAttr(Path path, String name) throws IOException {
      throw new NotInMountpointException(path, "getXAttr");
    }

    @Override
    public Map<String, byte[]> getXAttrs(Path path) throws IOException {
      throw new NotInMountpointException(path, "getXAttrs");
    }

    @Override
    public Map<String, byte[]> getXAttrs(Path path, List<String> names)
        throws IOException {
      throw new NotInMountpointException(path, "getXAttrs");
    }

    @Override
    public List<String> listXAttrs(Path path) throws IOException {
      throw new NotInMountpointException(path, "listXAttrs");
    }

    @Override
    public void removeXAttr(Path path, String name) throws IOException {
      checkPathIsSlash(path);
      throw readOnlyMountTable("removeXAttr", path);
    }

    @Override
    public Path createSnapshot(Path path, String snapshotName)
        throws IOException {
      checkPathIsSlash(path);
      throw readOnlyMountTable("createSnapshot", path);
    }

    @Override
    public void renameSnapshot(Path path, String snapshotOldName,
        String snapshotNewName) throws IOException {

View on GitHub (pinned to 2add963021)

Solutions

  1. Only call listXAttrs on paths inside a mount point
  2. Mount the queried prefix explicitly (fs.viewfs.mounttable.<cluster>.link.<path>)
  3. Configure linkFallback or linkMergeSlash so the root is backed by a real FileSystem
  4. Guard the tree walk with try/catch on NotInMountpointException and skip virtual dirs

Example fix

// before
List<String> names = fs.listXAttrs(dir);

// after
List<String> names;
try {
  names = fs.listXAttrs(dir);
} catch (NotInMountpointException e) {
  names = Collections.emptyList(); // virtual mount-table dir
}
Defensive patterns

Strategy: try-catch

Validate before calling

static boolean listXAttrsSafe(FileSystem fs, Path p, List<String> out) {
  try { out.addAll(fs.listXAttrs(p)); return true; }
  catch (NotInMountpointException e) { return false; } // virtual dir
}

Try / catch

try {
  names = fs.listXAttrs(dir);
} catch (NotInMountpointException e) {
  names = Collections.emptyList();
}

Prevention

When it happens

Trigger: fs.listXAttrs(new Path("/")) on viewfs://; listXAttrs on /project when the mount table only defines links under /project/a and /project/b; audit tools enumerating xattrs from the top of a viewfs namespace.

Common situations: Security/audit tooling (Ranger tag policies, Atlas) scanning federated namespaces through viewfs; scripts that walk from the root and list xattrs on every directory they traverse.

Related errors

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/8c29b796fcbb633e. Report an issue: GitHub.