apache/hadoop · error · YarnException

Can't find the queue by the given name: {}! Please check if

Error message

Can't find the queue by the given name: {}! Please check if queue {} is in the allocation file.

What it means

Thrown by SLSCapacityScheduler.getRealQueueName() when the SLS workload references a queue that getQueue(queue) cannot find, meaning no queue with that name exists in the loaded capacity-scheduler.xml configuration. SLS refuses to continue because it cannot map the simulated application to a real queue path.

Source

Thrown at hadoop-tools/hadoop-sls/src/main/java/org/apache/hadoop/yarn/sls/scheduler/SLSCapacityScheduler.java:133

      throw e;
    }
  }

  @Override
  public void propagatedHandle(SchedulerEvent schedulerEvent) {
    super.handle(schedulerEvent);
  }

  @Override
  public void serviceStop() throws Exception {
    schedulerCommons.stopMetrics();
    super.serviceStop();
  }


  public String getRealQueueName(String queue) throws YarnException {
    if (getQueue(queue) == null) {
      throw new YarnException("Can't find the queue by the given name: " + queue
          + "! Please check if queue " + queue + " is in the allocation file.");
    }
    return getQueue(queue).getQueuePath();
  }

  public SchedulerMetrics getSchedulerMetrics() {
    return schedulerCommons.getSchedulerMetrics();
  }

  @Override
  public Configuration getConf() {
    return conf;
  }

  public Tracker getTracker() {
    return schedulerCommons.getTracker();
  }

View on GitHub (pinned to 2add963021)

Solutions

1. Add the missing queue to capacity-scheduler.xml with valid capacity so it is initialized at scheduler startup
2. Use the exact queue name as configured (leaf queue name as CapacityScheduler sees it) in the "queue" field of every app in the SLS input JSON
3. Verify the SLS runner config actually loads the capacity-scheduler.xml you edited (check startup logs for parsed queues)
4. Cross-check queue names mechanically: list queues from the XML and diff against the set used in the workload file

Example fix

// before: sls input uses queue "etl" but allocation has root.default and root.prod
// capacity-scheduler.xml
<queue name="prod">
  <capacity>50</capacity>
</queue>
// after: add the queue and keep names in sync
<queue name="etl">
  <capacity>50</capacity>
  <maximum-capacity>100</maximum-capacity>
</queue>

// before SLSRunner start: parse capacity-scheduler.xml and diff against workload queues
Set<String> configured = new HashSet<>();
Document xml = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(new File("capacity-scheduler.xml"));
NodeList qs = xml.getElementsByTagName("queue");
for (int i = 0; i < qs.getLength(); i++) configured.add(qs.item(i).getAttributes().getNamedItem("name").getNodeValue());
Set<String> used = /* read "queue" fields from sls input json */;
used.removeAll(configured);
if (!used.isEmpty()) throw new IllegalStateException("Queues missing from allocation file: " + used);

wrap scheduler init/first getRealQueueName calls in try { ... } catch (YarnException e) { log e.getMessage(); } — the message already names the missing queue; fail fast with a config-fix hint.

When it happens

Trigger: Running SLSRunner with a capacity-scheduler setup where an app/task in the input JSON has a "queue" field that does not match any queue in the capacity-scheduler.xml pointed to by the SLS runner configuration (e.g. yarn.sls.scheduler.capacity.node-config / scheduler XML path).

Common situations: Using a fully-qualified path ('root.prod') where only the leaf name ('prod') is configured (or vice versa), queues renamed in capacity-scheduler.xml but not in the workload file, the wrong allocation file picked up because the SLS config points elsewhere, or a new queue added to the workload but never given capacity in the XML.

Related errors

• Can't find the queue by the given name: {}! Please check if
• Duration of a task shouldn't be less or equal to 0!
• ${diskValidator} DiskValidator class not found.
• dir + " is not a directory!"
• "Not creating intermediate history logDir: [" + doneDirPath

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/e81b912eaaf943c3. Report an issue: GitHub.