apache/hadoop · error · YarnRuntimeException

Invalid blacklistDisablePercent: {}. Should be an integer be

Error message

Invalid blacklistDisablePercent: {}. Should be an integer between 0 and 100 or -1 to disabled

What it means

RMContainerRequestor.serviceInit validates yarn.app.mapreduce.am.job.node-blacklisting.ignore-threshold-node-percent (default 33) and throws YarnRuntimeException when the value is outside [-1, 100]. -1 disables the threshold, 0..100 sets the percentage of blacklisted nodes at which blacklisting is ignored; anything else is meaningless and fails AM initialization immediately.

Source

Thrown at hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMContainerRequestor.java:190

      return sb.toString();
    }
  }

  @Override
  protected void serviceInit(Configuration conf) throws Exception {
    super.serviceInit(conf);
    nodeBlacklistingEnabled = 
      conf.getBoolean(MRJobConfig.MR_AM_JOB_NODE_BLACKLISTING_ENABLE, true);
    LOG.info("nodeBlacklistingEnabled:" + nodeBlacklistingEnabled);
    maxTaskFailuresPerNode = 
      conf.getInt(MRJobConfig.MAX_TASK_FAILURES_PER_TRACKER, 3);
    blacklistDisablePercent =
        conf.getInt(
            MRJobConfig.MR_AM_IGNORE_BLACKLISTING_BLACKLISTED_NODE_PERECENT,
            MRJobConfig.DEFAULT_MR_AM_IGNORE_BLACKLISTING_BLACKLISTED_NODE_PERCENT);
    LOG.info("maxTaskFailuresPerNode is " + maxTaskFailuresPerNode);
    if (blacklistDisablePercent < -1 || blacklistDisablePercent > 100) {
      throw new YarnRuntimeException("Invalid blacklistDisablePercent: "
          + blacklistDisablePercent
          + ". Should be an integer between 0 and 100 or -1 to disabled");
    }
    LOG.info("blacklistDisablePercent is " + blacklistDisablePercent);
  }

  protected AllocateResponse makeRemoteRequest() throws YarnException,
      IOException {
    applyRequestLimits();
    ResourceBlacklistRequest blacklistRequest =
        ResourceBlacklistRequest.newInstance(new ArrayList<String>(blacklistAdditions),
            new ArrayList<String>(blacklistRemovals));
    AllocateRequest allocateRequest =
        AllocateRequest.newInstance(lastResponseID,
          super.getApplicationProgress(), new ArrayList<ResourceRequest>(ask),
          new ArrayList<ContainerId>(release), blacklistRequest);
    AllocateResponse allocateResponse = scheduler.allocate(allocateRequest);
    lastResponseID = allocateResponse.getResponseId();

View on GitHub (pinned to 2add963021)

Solutions

1. Set the property to -1, 0, or a value in [0,100] — e.g. 33 (default) or -1 to disable the ignore threshold
2. If generated programmatically, clamp the computed integer before conf.set
3. Check the AM launch log: the exception text echoes the offending value read from config

Example fix

<!-- before -->
<property><name>yarn.app.mapreduce.am.job.node-blacklisting.ignore-threshold-node-percent</name><value>150</value></property>

<!-- after -->
<property><name>yarn.app.mapreduce.am.job.node-blacklisting.ignore-threshold-node-percent</name><value>33</value></property>

// clamp/validate before the AM ever sees it
int pct = conf.getInt(
    "yarn.app.mapreduce.am.job.node-blacklisting.ignore-threshold-node-percent", 33);
if (pct < -1 || pct > 100) {
  throw new IllegalArgumentException("blacklistDisablePercent must be -1 or 0..100, got " + pct);
}

When it happens

Trigger: Job/cluster config sets the percent key to e.g. 150 or -2; a conf.set call with a computed value that can go negative; typo like '33%' (NumberFormatException surfaces earlier) or an off-by-one in generated configs.

Common situations: Tuning node-blacklisting behavior from templates that mix up 'percent' and 'count'; automation writing 0-1 fractions (0.33) or >100 values; overriding MR_AM_IGNORE_BLACKLISTING_BLACKLISTED_NODE_PERECENT per job without bounds checking.

Related errors

• Improper queue name : {nameValue}
• Only 1 or 2 algorithm version is supported
• value cannot be blank
• ${value} is not in expected format.Expected format is <numbe
• Bad configuration of hadoop.security.key.provider.path at ${

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/c46256d8ccc3fe86. Report an issue: GitHub.