apache/hadoop · error · IOException

Incompatible event log version: ${version}

Error message

Incompatible event log version: ${version}

What it means

MapReduce job history files (.jhist) begin with a version line written by EventWriter: "Avro-Json" or "Avro-Binary". EventReader accepts exactly those two tags and throws IOException("Incompatible event log version: v") for any other first line, meaning the stream is not a job history in a known format.

Source

Thrown at hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapreduce/jobhistory/EventReader.java:82

   */
  @SuppressWarnings("deprecation")
  public EventReader(DataInputStream in) throws IOException {
    this.in = in;
    this.version = in.readLine();

    Schema myschema = new SpecificData(Event.class.getClassLoader()).getSchema(Event.class);
    Schema.Parser parser = new Schema.Parser();
    String eventschema = in.readLine();
    if (null != eventschema) {
      try {
        this.schema = parser.parse(eventschema);
        this.reader = new SpecificDatumReader(schema, myschema);
        if (EventWriter.VERSION.equals(version)) {
          this.decoder = DecoderFactory.get().jsonDecoder(schema, in);
        } else if (EventWriter.VERSION_BINARY.equals(version)) {
          this.decoder = DecoderFactory.get().binaryDecoder(in, null);
        } else {
          throw new IOException("Incompatible event log version: " + version);
        }
      } catch (AvroRuntimeException e) {
        throw new IOException(e);
      }
    } else {
      throw new IOException("Event schema string not parsed since its null");
    }
  }
  
  /**
   * Get the next event from the stream
   * @return the next event
   * @throws IOException
   */
  @SuppressWarnings("unchecked")
  public HistoryEvent getNextEvent() throws IOException {
    Event wrapper;
    try {

View on GitHub (pinned to 2add963021)

Solutions

1. Point history tooling at real .jhist files (done/ or intermediatedone/ history directories: job_*.jhist).
2. Verify the header before parsing: read the first line and check it is Avro-Json or Avro-Binary.
3. If the file was written by a different Hadoop line, read it with a Hadoop build matching the writer.
4. Discard files with unrecognized headers when scanning directories instead of failing the scan.

Example fix

# before: wrong file fed to history tooling
hadoop job -history /user/me/Logs/application_1690000000000_0001/job_1690000000000_0001-1690000000000-user-word+count-1-0-conf.xml

# after: real history file (check header first)
hdfs dfs -cat <path>/job_...jhist | head -1   # must print Avro-Json (or Avro-Binary)
hadoop job -history <path>/job_...jhist

// peek the header before building an EventReader
try (FSDataInputStream in = fs.open(path)) {
  String version = in.readLine();
  if (!("Avro-Json".equals(version) || "Avro-Binary".equals(version))) {
    // not a readable jhist - skip it
  }
}

catch (IOException e) around new EventReader(in): on "Incompatible event log version" close the stream, skip the file in directory scans, and report the file plus its first line for diagnosis.

When it happens

Trigger: Constructing new EventReader(inputStream) on a file whose first line is neither "Avro-Json" nor "Avro-Binary" - a non-history file (job.xml, logs, configs) passed to history parsing, a history written by an incompatible/renamed format, or a file whose first line was corrupted.

Common situations: Running 'mapreduce job -history' or HistoryViewer/JobHistoryParser against the wrong file (e.g. from the wrong directory or a job conf instead of .jhist); tooling that globs a directory and feeds every file to EventReader; cross-version history migration.

Related errors

• Counters version mismatch, expected ${groupFactory.version()
• Event schema string not parsed since its null
• Unknown mode: ${mode}
• Unable to initialize History Viewer
• Failure parsing JSON document: ${je.getMessage()}

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/f5639200a2f605b3. Report an issue: GitHub.