apache/hadoop · error · IOException

Failed to copy {srcFile} to {destFile} due to failure in Nat

Error message

Failed to copy {srcFile} to {destFile} due to failure in NativeIO#copyFileUnbuffered(). {e}

What it means

Storage.nativeCopyFileUnbuffered calls NativeIO.copyFileUnbuffered (sendfile/copy_file_range on Linux, CopyFile on Windows); a NativeIOException from that syscall is rewrapped here with both canonical paths and the original error text. The errno embedded in the message is the real diagnosis: ENOSPC (disk full), EIO (device error), EACCES, or ENOENT if the source vanished mid-copy.

Source

Thrown at hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/common/Storage.java:1363

        throw new IOException("Destination '" + parentFile
            + "' directory cannot be created");
      }
    }
    if (destFile.exists()) {
      if (FileUtil.canWrite(destFile) == false) {
        throw new IOException("Destination '" + destFile
            + "' exists but is read-only");
      } else {
        if (destFile.delete() == false) {
          throw new IOException("Destination '" + destFile
              + "' exists but cannot be deleted");
        }
      }
    }
    try {
      NativeIO.copyFileUnbuffered(srcFile, destFile);
    } catch (NativeIOException e) {
      throw new IOException("Failed to copy " + srcFile.getCanonicalPath()
          + " to " + destFile.getCanonicalPath()
          + " due to failure in NativeIO#copyFileUnbuffered(). "
          + e.toString());
    }
    if (srcFile.length() != destFile.length()) {
      throw new IOException("Failed to copy full contents from '" + srcFile
          + "' to '" + destFile + "'");
    }
    if (preserveFileDate) {
      if (destFile.setLastModified(srcFile.lastModified()) == false) {
        LOG.debug("Failed to preserve last modified date from'{}' to '{}'",
            srcFile, destFile);
      }
    }
  }

  /**
   * Recursively delete all the content of the directory first and then

View on GitHub (pinned to 2add963021)

Solutions

Read the errno in the message: ENOSPC -> free space or enlarge the volume; EIO -> check dmesg/RAID health
Retry the copy after resolving the transient cause (space freed, source stabilized)
Ensure no concurrent process deletes or chmods the source during checkpointing

Defensive patterns

Strategy: retry

Validate before calling

long free = dst.getParentFile().getUsableSpace();
if (free > 0 && free < srcFile.length()) {
  throw new IOException("insufficient space for copy: need " + srcFile.length() + ", have " + free);
}

Try / catch

for (int i = 0; i < 3; i++) {
  try {
    Storage.nativeCopyFileUnbuffered(src, dst, true);
    break;
  } catch (IOException e) {
    if (!isTransient(e.getMessage())) throw e; // ENOSPC/EACCES are not transient
    sleepBackoff(i);
  }
}

Prevention

Monitor free space on name/checkpoint volumes; ENOSPC (errno 28) is the most common cause
Retry only genuinely transient errors (EIO on flaky hardware, brief races), never permission errors
Keep antivirus and backup tools from holding storage files during checkpoints

When it happens

Trigger: Destination disk full (errno 28/ENOSPC); failing disk or controller (EIO); source removed by a concurrent process during the copy; permission revoked between the pre-checks and the syscall.

Common situations: Checkpoint image copies filling a small name-dir volume; dying disks during fsimage transfer; concurrent cleanup scripts deleting in-flight files.

Related errors

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/9badf51b11f092a0. Report an issue: GitHub.