[BUG] org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests.testQueryFiltering test failure

[kotwanikunal]

Describe the bug
A test failure, possibly flakey - Noticed with #4786

To Reproduce

REPRODUCE WITH: ./gradlew ':server:test' --tests "org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests.testQueryFiltering" -Dtests.seed=31AC51DE66176DE -Dtests.security.manager=true -Dtests.jvm.argline="-XX:TieredStopAtLevel=1 -XX:ReservedCodeCacheSize=64m" -Dtests.locale=nl-BE -Dtests.timezone=America/Boa_Vista -Druntime.java=17

Expected behavior

• Test should pass.

Plugins

• None

Screenshots

org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests > testQueryFiltering FAILED
    java.lang.AssertionError: 
    Expected: within <10.0> percent of <125.0>
         but: was <111.1025641025641>
        at __randomizedtesting.SeedInfo.seed([31AC51DE66176DE:C77329D1892D88C7]:0)
        at org.hamcrest.MatcherAssert.assertThat(MatcherAssert.java:18)
        at org.junit.Assert.assertThat(Assert.java:964)
        at org.junit.Assert.assertThat(Assert.java:930)
        at org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests.lambda$testQueryFiltering$11(MedianAbsoluteDeviationAggregatorTests.java:154)
        at org.opensearch.search.aggregations.AggregatorTestCase.testCase(AggregatorTestCase.java:619)
        at org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests.testAggregation(MedianAbsoluteDeviationAggregatorTests.java:255)
        at org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests.testAggregation(MedianAbsoluteDeviationAggregatorTests.java:245)
        at org.opensearch.search.aggregations.metrics.MedianAbsoluteDeviationAggregatorTests.testQueryFiltering(MedianAbsoluteDeviationAggregatorTests.java:149)

Host/Environment (please complete the following information):

• Jenkins build/check environment

Additional context
N/A

[andrross]

A quick investigation suggests that when this random call generates a value less than 21 for the random double, then the test will fail.

[SohamRatnaparkhi]

Hey @andrross, I would like to work on this issue. Can you assign it to me?

[andrross]

Hey @andrross, I would like to work on this issue. Can you assign it to me?

Done! Thanks for looking into this!

[SohamRatnaparkhi]

Can anyone guide me in short as to what I have to do to solve this issue?
Basically, how can I start; where are the required files wherein I have to make changes.

[andrross]

@SohamRatnaparkhi So changing the 20 to a 21 in this test case would likely fix the test failure, but it would be better (and is much harder :) ) to answer the question of why a value below 21 causes the test to fail. It could be that the original test author just miscalculated a bit, or it could be a bug. Answering this question requires digging into exactly what the test is intending to do.

[SohamRatnaparkhi]

Okay, I will try and find out the reason too along with changing 20 to a 21.

[SohamRatnaparkhi]

Hi, tried searching in the code base as to why code breaks at value = 20. Anyways, So I am making a PR by changing 20 to 21.

[austintlee]

I can get it to fail with 21 and 22. Any progress on this? Can I take this on?

[austintlee]

@andrross @SohamRatnaparkhi can you assign this one to me?

[andrross]

@austintlee For some reason GitHub does not allow me to add you in the assignee field, but please feel free to take this on. Thanks!

[dblock]

Closed via #7317