Historical errors when using PeriodGranity in the groupBy query

[soullkk]

Historical errors when using PeriodGranity in the groupBy query

Affected Version

Apache Druid 28.0.1.

Description

Please include as much detailed information about the problem as possible.

• Steps to reproduce the problem
  1、Create a datasource(like ODAEDATASET__DEFAULT_lkk_person_olap__DEFAULT) and ingest data with timestamps greater than 2024-11-18T12:00:00 and less than 2024-11-18T13:00:00
  2、Use the "/druid/v2" to perform the following query :

{
    "queryType": "groupBy",
    "dataSource": "ODAEDATASET__DEFAULT_lkk_person_olap__DEFAULT",
    "granularity": {
        "type": "period",
        "period": "PT5H",
        "timeZone": "Asia/Singapore"
    },
    "dimensions": [{
            "type": "default",
            "dimension": "address",
            "outputName": "address",
            "outputType": "STRING"
        }
    ],
    "limitSpec": {
        "type": "default",
        "limit": 14400,
        "columns": []
    },
    "intervals": ["2024-11-18T12:05:00.000Z/2024-11-18T12:15:00.000Z"],
    "context": {
        "timeout": 30000,
        "maxRowsQueuedForOrdering": 128575
    }
}

• The error message or stack traces encountered. Providing more context, such as nearby log messages or even entire logs, can be helpful.
  The error stack of the historical node is as follows :

2024-11-19 03:21:31,097 ERROR [processing-4][][org.apache.druid.query.groupby.epinephelinae.GroupByMergingQueryRunnerV2] Exception with one of the sequences!
java.lang.IllegalArgumentException: The end instant must be greater than the start instant
	at org.joda.time.base.AbstractInterval.checkInterval(AbstractInterval.java:63) ~[joda-time-2.12.5.jar:2.12.5]
	at org.joda.time.base.BaseInterval.<init>(BaseInterval.java:94) ~[joda-time-2.12.5.jar:2.12.5]
	at org.joda.time.Interval.<init>(Interval.java:201) ~[joda-time-2.12.5.jar:2.12.5]
	at org.apache.druid.query.vector.VectorCursorGranularizer.create(VectorCursorGranularizer.java:82) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.groupby.epinephelinae.vector.VectorGroupByEngine$VectorGroupByEngineIterator.<init>(VectorGroupByEngine.java:286) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.groupby.epinephelinae.vector.VectorGroupByEngine$1.make(VectorGroupByEngine.java:192) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.groupby.epinephelinae.vector.VectorGroupByEngine$1.make(VectorGroupByEngine.java:144) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.BaseSequence.accumulate(BaseSequence.java:39) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.SequenceWrapper.wrap(SequenceWrapper.java:55) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.SequenceWrapper.wrap(SequenceWrapper.java:55) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.LazySequence.accumulate(LazySequence.java:40) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.SequenceWrapper.wrap(SequenceWrapper.java:55) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.MappedSequence.accumulate(MappedSequence.java:43) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.SequenceWrapper.wrap(SequenceWrapper.java:55) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.LazySequence.accumulate(LazySequence.java:40) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.SequenceWrapper.wrap(SequenceWrapper.java:55) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.spec.SpecificSegmentQueryRunner$1.accumulate(SpecificSegmentQueryRunner.java:98) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.spec.SpecificSegmentQueryRunner.doNamed(SpecificSegmentQueryRunner.java:185) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.spec.SpecificSegmentQueryRunner.access$100(SpecificSegmentQueryRunner.java:44) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.spec.SpecificSegmentQueryRunner$2.wrap(SpecificSegmentQueryRunner.java:165) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence$1.get(WrappingSequence.java:50) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.CPUTimeMetricQueryRunner$1.wrap(CPUTimeMetricQueryRunner.java:77) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.java.util.common.guava.WrappingSequence.accumulate(WrappingSequence.java:45) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.groupby.epinephelinae.GroupByMergingQueryRunnerV2$1$1$1.call(GroupByMergingQueryRunnerV2.java:252) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at org.apache.druid.query.groupby.epinephelinae.GroupByMergingQueryRunnerV2$1$1$1.call(GroupByMergingQueryRunnerV2.java:239) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at java.util.concurrent.FutureTask.run(FutureTask.java:266) ~[?:1.8.0_432]
	at org.apache.druid.query.PrioritizedListenableFutureTask.run(PrioritizedExecutorService.java:259) ~[druid-processing-28.0.1-htrunk15.jar:?]
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) ~[?:1.8.0_432]
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) ~[?:1.8.0_432]
	at java.lang.Thread.run(Thread.java:750) ~[?:1.8.0_432]

• Any debugging that you have already done
  If the interval of a segment is like this：2024-11-18T12:10:00.000 --- 2024-11-18T12:15:00.000
  The result of granularity.bucketEnd(maxTime) will be less than minTime when using the query mentioned earlier.

  @Nullable
  public static VectorCursorGranularizer create(
      final StorageAdapter storageAdapter,
      final VectorCursor cursor,
      final Granularity granularity,
      final Interval queryInterval
  )
  {
    final DateTime minTime = storageAdapter.getMinTime();
    final DateTime maxTime = storageAdapter.getMaxTime();

    final Interval storageAdapterInterval = new Interval(minTime, granularity.bucketEnd(maxTime));

# PeriodGranularity.truncate(long t):
# period = "PT5H"
# t = 2024-11-18T12:15:00.000
# timeZone = "Asia/Singapore"
# origin = -27000000

    final int hours = period.getHours();
    if (hours > 0) {
      if (hours > 1 || hasOrigin) {
        // align on multiples from origin
        long h = chronology.hours().getDifferenceAsLong(t, origin);
        h -= h % hours;
        long tt = chronology.hours().add(origin, h);
        // always round down to the previous period (for timestamps prior to origin)
        if (t < tt && origin > 0) {
          t = chronology.hours().add(tt, -hours);
        } else if (t > tt && origin < 0) {
          t = chronology.minuteOfHour().roundFloor(tt);
          t = chronology.minuteOfHour().set(t, 0);
        } else {
          t = tt;
        }
        return t;
      } else {
        return chronology.hourOfDay().roundFloor(t);
      }
    }

The result of "long tt = chronology.hours().add(origin, h)" will be "2024-11-18T07:30:00.000Z"
"t > tt && origin < 0 " is valid.
so "t = chronology.minuteOfHour().set(t, 0)" will be executed
the return will be t = "2024-11-18T07:00:00.000Z"
so granularity.bucketEnd(maxTime) = "2024-11-18T12:00:00.000Z", it is less than 2024-11-18T12:10:00.000 (the start time of interval)

[soullkk]

This issue may be related to #4073