[CI] IndicesClientIT.testDataStreams failure

[dimitris-athanasiou]

Build scan: https://gradle-enterprise.elastic.co/s/rtz2uf3u5fuwm

Repro line: ./gradlew ':client:rest-high-level:asyncIntegTestRunner' --tests "org.elasticsearch.client.IndicesClientIT.testDataStreams" -Dtests.seed=E89C70116AD44583 -Dtests.security.manager=true -Dtests.locale=ja-JP-u-ca-japanese-x-lvariant-JP -Dtests.timezone=Australia/LHI -Druntime.java=8

Reproduces locally?: No

Applicable branches: 7.9

Failure history: First time

Failure excerpt:

org.elasticsearch.client.IndicesClientIT > testDataStreams FAILED
    java.lang.AssertionError: 
    Expected: not <0L>
         but: was <0L>
        at __randomizedtesting.SeedInfo.seed([E89C70116AD44583:AE38E9FD1394286D]:0)
        at org.hamcrest.MatcherAssert.assertThat(MatcherAssert.java:18)
        at org.junit.Assert.assertThat(Assert.java:956)
        at org.junit.Assert.assertThat(Assert.java:923)
        at org.elasticsearch.client.IndicesClientIT.testDataStreams(IndicesClientIT.java:2084)

[elasticmachine]

Pinging @elastic/es-core-features (:Core/Features/Data streams)

[martijnvg]

The test fails, because the total store size is expected to be not zero, but it is zero.
I think this can happen, because no shards of the backing index has been allocated,
which is possible, because we don't wait for shards to get allocated. Just some checks
that are based on things in the cluster state. I think we should wait for at least
on shard to get allocated of the backing index.

[nik9000]

@martijnvg this seems to have come back. 🧟?