apache/hadoop · error · IOException

Could not initialize target File System for URI : {targetDir

Error message

Could not initialize target File System for URI : {targetDirLinkList[0]}

What it means

In InodeTree, each single mount link holds a target FileSystem that is created lazily: the first access to the mounted path calls fileSystemInitMethod.apply(URI.create(target)). If that returns null, IOException('Could not initialize target File System for URI : <uri>') is thrown. The failure surfaces at first path access, not when the ViewFileSystem is created.

Source

Thrown at hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/viewfs/InodeTree.java:408

     * Get the instance of FileSystem to use, creating one if needed.
     * @return An Initialized instance of T
     * @throws IOException raised on errors performing I/O.
     */
    public T getTargetFileSystem() throws IOException {
      if (targetFileSystem != null) {
        return targetFileSystem;
      }
      // For non NFLY and MERGE links, we initialize the FileSystem when the
      // corresponding mount path is accessed.
      if (targetDirLinkList.length == 1) {
        synchronized (lock) {
          if (targetFileSystem != null) {
            return targetFileSystem;
          }
          targetFileSystem =
              fileSystemInitMethod.apply(URI.create(targetDirLinkList[0]));
          if (targetFileSystem == null) {
            throw new IOException(
                "Could not initialize target File System for URI : " +
                    targetDirLinkList[0]);
          }
        }
      }
      return targetFileSystem;
    }

    T getTargetFileSystemForClose() throws IOException {
      return targetFileSystem;
    }

  }

  private void createLink(final String src, final String target,
      final LinkType linkType, final String settings,
      final UserGroupInformation aUgi,
      final Configuration config)

View on GitHub (pinned to 2add963021)

Solutions

1. Verify the target standalone first: FileSystem.get(URI.create(linkTarget), conf) must succeed
2. Fix the scheme/URI in the fs.viewfs.mounttable.<name>.link.<path> entry
3. Ensure the FileSystem implementation jar is on the client classpath

Example fix

<!-- before -->
<property><name>fs.viewfs.mounttable.cluster.link./data</name>
  <value>hdfs2://nnA/data</value></property>

<!-- after -->
<property><name>fs.viewfs.mounttable.cluster.link./data</name>
  <value>hdfs://nnA/data</value></property>

static boolean targetResolvable(String linkTarget, Configuration conf) {
  try {
    return FileSystem.get(URI.create(linkTarget), conf) != null;
  } catch (IOException e) {
    return false;
  }
}

Wrap the first access to each mount point in try/catch IOException; on 'Could not initialize target File System for URI', extract the URI from the message and verify it standalone with FileSystem.get(uri, conf).

When it happens

Trigger: A mount table link whose target URI uses a scheme with no FileSystem implementation or a broken factory; accessing any path under such a mount; custom fileSystemInitMethod implementations that return null.

Common situations: Typo'd scheme in a link target ('hdfs2://', 'hdf://'); missing connector jar for the target store; federated configs referencing decommissioned clusters.

Related errors

• Uri without authority: {uri}
• Wrong FS: {path}, expected: {this.getUri()}
• Wrong FS: {path} and port={thatPort}, expected: {this.getUri
• ViewFs: Non absolute mount name in config:{src}
• Path {nextInode.fullPath} already exists as link

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/bd1a713aecd48e8e. Report an issue: GitHub.