apache/hadoop · error · IOException

Cannot start trash emptier with negative interval. Set fs.tr

Error message

Cannot start trash emptier with negative interval. Set fs.trash.interval to a positive value.

What it means

NameNode.startTrashEmptier reads fs.trash.interval from the configuration: 0 disables the emptier (method returns), but any negative value throws IOException at startup because a trash-emptier thread cannot be scheduled with a negative interval. The NameNode refuses to start the background emptier with a nonsensical period.

Source

Thrown at hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/NameNode.java:1066

    if (plugins != null) {
      for (ServicePlugin p : plugins) {
        try {
          p.stop();
        } catch (Throwable t) {
          LOG.warn("ServicePlugin " + p + " could not be stopped", t);
        }
      }
    }   
    stopHttpServer();
  }
  
  private void startTrashEmptier(final Configuration conf) throws IOException {
    long trashInterval =
        conf.getLong(FS_TRASH_INTERVAL_KEY, FS_TRASH_INTERVAL_DEFAULT);
    if (trashInterval == 0) {
      return;
    } else if (trashInterval < 0) {
      throw new IOException("Cannot start trash emptier with negative interval."
          + " Set " + FS_TRASH_INTERVAL_KEY + " to a positive value.");
    }
    
    // This may be called from the transitionToActive code path, in which
    // case the current user is the administrator, not the NN. The trash
    // emptier needs to run as the NN. See HDFS-3972.
    FileSystem fs = SecurityUtil.doAsLoginUser(
        new PrivilegedExceptionAction<FileSystem>() {
          @Override
          public FileSystem run() throws IOException {
            FileSystem dfs = new DistributedFileSystem();
            dfs.initialize(FileSystem.getDefaultUri(conf), conf);
            return dfs;
          }
        });
    this.emptier = new Thread(new Trash(fs, conf).getEmptier(), "Trash Emptier");
    this.emptier.setDaemon(true);
    this.emptier.start();

View on GitHub (pinned to 2add963021)

Solutions

1. Set fs.trash.interval to 0 in core-site.xml to disable the trash emptier, or to a positive value in minutes (e.g., 60, 1440).
2. Restart the NameNode after fixing the configuration so startTrashEmptier re-reads it.
3. Fix the automation/template that produced the negative value so it can never emit one again.

Example fix

<!-- before: core-site.xml -->
<property><name>fs.trash.interval</name><value>-1</value></property>

<!-- after -->
<property><name>fs.trash.interval</name><value>0</value></property> <!-- 0 disables trash emptier; positive = retention minutes -->

// fail fast at deploy time, before the NN hits it at startup
long interval = conf.getLong(FS_TRASH_INTERVAL_KEY, FS_TRASH_INTERVAL_DEFAULT);
Preconditions.checkArgument(interval >= 0,
    "fs.trash.interval must be >= 0 (0 disables), got %s", interval);

When it happens

Trigger: core-site.xml (or an equivalent config source) sets fs.trash.interval to a negative number and the NameNode starts or transitions to active (startTrashEmptier runs on the active NN; see also HDFS-3972 where it runs as the login user).

Common situations: Operators set -1 trying to 'disable' trash (the correct disable value is 0); configuration management templates that compute the interval and can emit negatives; copy-paste from docs of other systems where -1 means infinite.

Related errors

• Unknown nameservice: {}
• Configuration has multiple addresses that match local node's
• Configuration dfs.namenode.rpc-address must be suffixed with
• Incompatible node types: storageType={storageType} but Stora
• All specified directories are not accessible or do not exist

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/fd2d65ba8fe6dccf. Report an issue: GitHub.