[fix][test] Fix flaky AdminApiOffloadTest.testOffload status race

[merlimat]

Motivation

AdminApiOffloadTest.testOffload (parameterized as testOffloadV1 and testOffloadV2) checks the offload status immediately after completing the offload promise exceptionally:

promise.completeExceptionally(new Exception("Some random failure"));

assertEquals(admin.topics().offloadStatus(topicName).getStatus(),
                    LongRunningProcessStatus.Status.ERROR);

The failure handler that updates the offload status to ERROR runs asynchronously after the promise completes. The synchronous assertion races with that handler and occasionally observes RUNNING, producing the flake expected [ERROR] but found [RUNNING].

Example failure: https://scans.gradle.com/s/nv4xn5dwytjvs/tests/task/:pulsar-broker:test/details/org.apache.pulsar.broker.admin.AdminApiOffloadTest/testOffloadV2/2/output

Modifications

Wrap the ERROR status assertion in Awaitility.untilAsserted(...) so the test waits for the failure handler to update the status, mirroring the success path immediately below it (which already used Awaitility).

Verifying this change

This change is already covered by existing tests:

• AdminApiOffloadTest.testOffloadV1
• AdminApiOffloadTest.testOffloadV2

Locally I ran both 3 times in a row with fresh Gradle daemons; each run took ~0.4–0.9s and all passed.

Does this pull request potentially affect one of the following parts:

• Dependencies (add or upgrade a dependency)
• The public API
• The schema
• The default values of configurations
• The threading model
• The binary protocol
• The REST endpoints
• The admin CLI options
• The metrics
• Anything that affects deployment

[lhotari]

LGTM