apache/hadoop · error · ExitUtil.ExitException

42

42

Error message

Filesystem is not S3A URL: {fsURI}

What it means

The 'bucket' subcommand treats its single non-option argument as the URL of the bucket to create: it builds a Path, takes its URI and uses the host as the bucket name. If the URI scheme is not exactly 's3a', it exits with code 42 (EXIT_USAGE) and 'Filesystem is not S3A URL: <uri>'.

Source

Thrown at hadoop-tools/hadoop-aws/src/main/java/org/apache/hadoop/fs/s3a/tools/BucketTool.java:179

    LOG.debug("Supplied arguments: {}", String.join(", ", args));
    final List<String> parsedArgs = parseArgsWithErrorReporting(args);

    CommandFormat command = getCommandFormat();
    boolean create = command.getOpt(CREATE);

    Optional<String> endpoint = getOptionalString(OPT_ENDPOINT);
    Optional<String> region = getOptionalString(OPT_REGION);
    Optional<String> zone = getOptionalString(OPT_ZONE);


    final String bucketPath = parsedArgs.get(0);
    final Path source = new Path(bucketPath);
    URI fsURI = source.toUri();
    String bucket = fsURI.getHost();

    println(out, "Filesystem %s", fsURI);
    if (!"s3a".equals(fsURI.getScheme())) {
      throw new ExitUtil.ExitException(EXIT_USAGE, "Filesystem is not S3A URL: " + fsURI);
    }
    println(out, "Options region=%s endpoint=%s zone=%s s3a://%s",
        region.orElse("(unset)"),
        endpoint.orElse("(unset)"),
        zone.orElse("(unset)"),
        bucket);
    if (!create) {
      errorln(getUsage());
      println(out, "Supplied arguments: ["
          + String.join(", ", parsedArgs)
          + "]");
      throw new ExitUtil.ExitException(EXIT_USAGE,
          "required option not found: -create");
    }

    final Configuration conf = getConf();

    removeBucketOverrides(bucket, conf,

View on GitHub (pinned to 2add963021)

Solutions

1. Pass the fully qualified URL with the s3a scheme: 'hadoop aws s3guard bucket -create s3a://my-bucket'
2. If your script holds a bare bucket name, prefix it: "s3a://$BUCKET"
3. Do not substitute s3:// for s3a:// - the S3A connector only recognizes its own scheme

Example fix

# before
hadoop aws s3guard bucket -create s3://my-bucket
# after
hadoop aws s3guard bucket -create s3a://my-bucket

URI u = new Path(bucketArg).toUri();
if (!"s3a".equals(u.getScheme()) || u.getHost() == null) {
  throw new IllegalArgumentException("bucket tool requires s3a://bucket URL, got: " + bucketArg);
}

try {
  int rc = new BucketTool(conf).exec(args);
} catch (ExitUtil.ExitException e) {
  if (e.getExitCode() == 42 && e.getMessage().contains("not S3A URL")) {
    // rewrite the argument to s3a:// form and retry once
  }
}

When it happens

Trigger: Passing 's3://bucket' (AWS CLI habit), a bare 'bucketname' (URI with null scheme and null host), 'wasb://...' or another scheme, or a local path like '/tmp/x'; any of these fails the 's3a'.equals(scheme) check.

Common situations: Muscle memory from the AWS CLI ('aws s3 ...' uses s3://); scripts that pass just the bucket name; operators mixing HDFS or ABFS URLs into an S3 tool invocation.

Related errors

• 42
• 42
• Bad argument: {}
• Expect either -getlevel or -setlevel
• "Unexpected argument " + args[nextArgIndex]

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/198eed1517985cff. Report an issue: GitHub.