apache/druid · error · IAE

Unable to parse timestamps with format

Error message

Unable to parse timestamps with format [%s]

What it means

TimestampParser.createTimestampParser builds a DateTimeFormatter from the configured format string; if the Joda-Time pattern is invalid or cannot be constructed, it throws IllegalArgumentException wrapping the cause with 'Unable to parse timestamps with format [<format>]'. This fails fast at parser construction, before any data is seen.

Solutions

1. Correct the format string to valid Joda-Time syntax (e.g. "yyyy-MM-dd'T'HH:mm:ss.SSSZ")
2. Verify each token against the Joda-Time DateTimeFormat documentation
3. Test the format with DateTimeFormat.forPattern(format) in isolation before wiring it into ingestion
4. If the data is actually numeric timestamps, use createNumericTimestampParser instead

Example fix

// before
TimestampParser.createTimestampParser("%Y-%m-%d %H:%M:%S") // strftime-style: throws
// after
TimestampParser.createTimestampParser("yyyy-MM-dd HH:mm:ss")

try {
  org.joda.time.format.DateTimeFormat.forPattern(format);
} catch (IllegalArgumentException e) {
  throw new IllegalArgumentException("Invalid timestamp format: " + format, e);
}

boolean isValidJodaFormat(String format) {
  try { org.joda.time.format.DateTimeFormat.forPattern(format); return true; }
  catch (Exception e) { return false; }
}

try {
  Function<Number, DateTime> p = TimestampParser.createTimestampParser(format);
} catch (IllegalArgumentException e) {
  throw new IllegalStateException("Fix timestamp format in config: " + format, e);
}

When it happens

Trigger: Calling createTimestampParser(String format) with a format string that Joda-Time's DateTimeFormat cannot parse (bad pattern letters, invalid literals).

Common situations: Typos in format tokens (e.g. 'YYY' vs 'yyyy', 'hh' vs 'HH'); copying SimpleDateFormat patterns that differ from Joda syntax; formats pasted from other systems like Python strftime; null or empty format handled earlier but invalid syntax reaches here.

Understand the failure class

• Parsing and encoding errors: unexpected token, malformed input — why parsers reject input and how to find the real culprit.

Related errors

• Adjacent intervals are not sorted
• Element of intervals is null
• Empty list of intervals
• Encountered row with timestamp
• Protobuf message could not be parsed

AI-assisted analysis of apache/druid@9b90983fd2 (2026-09-07). Data as JSON: /api/errors/ed48cb746a6d3eb7. Report an issue: GitHub.

Appendix: source

Thrown at processing/src/main/java/org/apache/druid/java/util/common/parsers/TimestampParser.java:97

        Preconditions.checkArgument(!Strings.isNullOrEmpty(input), "null timestamp");
        return numericFun.apply(Long.parseLong(ParserUtils.stripQuotes(input)));
      };
    } else if ("ruby".equalsIgnoreCase(format)) {
      final Function<Number, DateTime> numericFun = createNumericTimestampParser(format);
      return input -> {
        Preconditions.checkArgument(!Strings.isNullOrEmpty(input), "null timestamp");
        return numericFun.apply(Double.parseDouble(ParserUtils.stripQuotes(input)));
      };
    } else {
      try {
        final DateTimes.UtcFormatter formatter = DateTimes.wrapFormatter(DateTimeFormat.forPattern(format));
        return input -> {
          Preconditions.checkArgument(!Strings.isNullOrEmpty(input), "null timestamp");
          return formatter.parse(ParserUtils.stripQuotes(input));
        };
      }
      catch (Exception e) {
        throw new IAE(e, "Unable to parse timestamps with format [%s]", format);
      }
    }
  }

  public static Function<Number, DateTime> createNumericTimestampParser(
      final String format
  )
  {
    if ("posix".equalsIgnoreCase(format)) {
      return input -> DateTimes.utc(TimeUnit.SECONDS.toMillis(input.longValue()));
    } else if ("micro".equalsIgnoreCase(format)) {
      return input -> DateTimes.utc(TimeUnit.MICROSECONDS.toMillis(input.longValue()));
    } else if ("nano".equalsIgnoreCase(format)) {
      return input -> DateTimes.utc(TimeUnit.NANOSECONDS.toMillis(input.longValue()));
    } else if ("ruby".equalsIgnoreCase(format)) {
      return input -> DateTimes.utc(Double.valueOf(input.doubleValue() * 1000).longValue());
    } else {
      return input -> DateTimes.utc(input.longValue());

View on GitHub (pinned to 9b90983fd2)