apache/hadoop · error · YarnRuntimeException

Unknown task symbol: {}

Error message

Unknown task symbol: {}

What it means

MRApps.taskType(String) translates the one-letter task symbols used in MRv2 task/attempt IDs and web query parameters: only 'm' (MAP) and 'r' (REDUCE) are legal; anything else throws YarnRuntimeException. It is called from the MR AppMaster and JobHistory web UI/REST layer to parse the task-type request parameter, so the exception surfaces as a failed web request when the parameter is wrong.

Source

Thrown at hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapreduce/v2/util/MRApps.java:158

    PENDING(new TaskState[]{TaskState.SCHEDULED}),
    COMPLETED(new TaskState[]{TaskState.SUCCEEDED, TaskState.FAILED, TaskState.KILLED});

    private final List<TaskState> correspondingStates;

    private TaskStateUI(TaskState[] correspondingStates) {
      this.correspondingStates = Arrays.asList(correspondingStates);
    }

    public boolean correspondsTo(TaskState state) {
      return this.correspondingStates.contains(state);
    }
  }

  public static TaskType taskType(String symbol) {
    // JDK 7 supports switch on strings
    if (symbol.equals("m")) return TaskType.MAP;
    if (symbol.equals("r")) return TaskType.REDUCE;
    throw new YarnRuntimeException("Unknown task symbol: "+ symbol);
  }

  public static TaskAttemptStateUI taskAttemptState(String attemptStateStr) {
    return TaskAttemptStateUI.valueOf(attemptStateStr);
  }

  public static TaskStateUI taskState(String taskStateStr) {
    return TaskStateUI.valueOf(taskStateStr);
  }

  // gets the base name of the MapReduce framework or null if no
  // framework was configured
  private static String getMRFrameworkName(Configuration conf) {
    String frameworkName = null;
    String framework =
        conf.get(MRJobConfig.MAPREDUCE_APPLICATION_FRAMEWORK_PATH, "");
    if (!framework.isEmpty()) {
      URI uri;

View on GitHub (pinned to 2add963021)

Solutions

1. Use the one-letter symbols: type=m for maps, type=r for reduces
2. Omit the type parameter entirely to list all tasks
3. When scripting against the API, validate the param against ^[mr]$ before issuing the request

Example fix

# before
GET /ws/v1/mapreduce/jobs/job_1400_0001/tasks?type=map

# after
GET /ws/v1/mapreduce/jobs/job_1400_0001/tasks?type=m

static boolean isValidTaskSymbol(String s) {
  return "m".equals(s) || "r".equals(s);
}
// in REST clients / scrapers: validate ?type= before calling the MR web APIs

static boolean isValidTaskSymbol(String s) {
  return "m".equals(s) || "r".equals(s);
}

try {
  TaskType t = MRApps.taskType(typeParam);
} catch (YarnRuntimeException e) {
  response.setStatus(HttpServletResponse.SC_BAD_REQUEST);
  // "type must be 'm' or 'r'"
}

When it happens

Trigger: Hitting AM or JHS endpoints with an invalid type parameter, e.g. GET /proxy/application_1/ws/v1/mapreduce/jobs/job_1/tasks?type=map or ?type=setup (full words and setup/cleanup symbols are rejected); same for JSP pages like tasks.jsp?type=x.

Common situations: Scripts or monitoring dashboards scraping MR web UIs passing 'map'/'reduce' instead of 'm'/'r'; humans typing the full word into the URL; curl examples copied from other REST APIs (e.g. YARN's own, which spells types differently).

Related errors

• Unrecognized task type: {}
• startedTimeEnd must be greater than 0
• startedTimeEnd must be greater than startTimeBegin
• finishedTimeBegin must be greater than 0
• finishedTimeEnd must be greater than 0

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/4c25153cb2154a11. Report an issue: GitHub.