apache/hadoop · error · NotInMountpointException

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

Error message

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

What it means

ViewFileSystem.getQuotaUsage throws NotInMountpointException from InternalDirOfViewFs when the path resolves to an internal mount-table directory instead of a mounted FileSystem. Quota information lives on the target (per-cluster) FileSystem, so a virtual node cannot answer the query. The exception is unchecked (a UnsupportedOperationException subtype).

Source

Thrown at hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/viewfs/ViewFileSystem.java:1888

    }

    @Override
    public void renameSnapshot(Path path, String snapshotOldName,
        String snapshotNewName) throws IOException {
      checkPathIsSlash(path);
      throw readOnlyMountTable("renameSnapshot", path);
    }

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

    @Override
    public QuotaUsage getQuotaUsage(Path f) throws IOException {
      throw new NotInMountpointException(f, "getQuotaUsage");
    }

    @Override
    public void satisfyStoragePolicy(Path src) throws IOException {
      checkPathIsSlash(src);
      throw readOnlyMountTable("satisfyStoragePolicy", src);
    }

    @Override
    public void setStoragePolicy(Path src, String policyName)
        throws IOException {
      checkPathIsSlash(src);
      throw readOnlyMountTable("setStoragePolicy", src);
    }

    @Override
    public void unsetStoragePolicy(Path src) throws IOException {
      checkPathIsSlash(src);

View on GitHub (pinned to 2add963021)

Solutions

  1. Query quota per mount point: iterate getMountPoints() and call getQuotaUsage on each mounted path
  2. Add a link for the queried prefix, or set linkMergeSlash/linkFallback so / resolves to a cluster
  3. Catch NotInMountpointException and report quota as unavailable for that subtree

Example fix

// before
QuotaUsage q = vfs.getQuotaUsage(new Path("/")); // throws

// after
for (MountPoint mp : vfs.getMountPoints()) {
  QuotaUsage q = vfs.getQuotaUsage(mp.getMountedOnPath());
  // aggregate per-cluster quota here
}
Defensive patterns

Strategy: try-catch

Validate before calling

// Validate coverage before asking for quota
static boolean quotaQueryable(ViewFileSystem vfs, Path p) {
  return vfs.getMountPoints().stream()
      .anyMatch(mp -> p.equals(mp.getMountedOnPath())
          || p.toString().startsWith(mp.getMountedOnPath() + "/"));
}

Try / catch

try {
  QuotaUsage q = vfs.getQuotaUsage(p);
} catch (NotInMountpointException e) {
  // aggregate per mount point instead
}

Prevention

When it happens

Trigger: fs.getQuotaUsage(new Path("/")) or getQuotaUsage on a virtual ancestor dir (e.g. /data with only /data/ds1 mounted) via a viewfs:// URI; capacity dashboards polling the federated root.

Common situations: Federation capacity-reporting tools, `hdfs dfsadmin`-style monitoring, or Hive/Spark code calling getQuotaUsage with the cluster defaultFS = viewfs://; the mount table lacking a root-level link.

Related errors

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/cacedaad6151f4e4. Report an issue: GitHub.