[fix][test] Fix flaky OneWayReplicatorDeduplicationTest.testDeduplication

[Praveenkumar76]

Fixes apache#25141

Motivation

OneWayReplicatorDeduplicationTest.testDeduplication is flaky because MessageDeduplication.takeSnapshot is designed to drop concurrent snapshot requests if an existing snapshot operation is already in progress. The test relies on specific positions being persisted, but if a manual trigger is dropped due to this intentional "lossy" design, the test fails with a ConditionTimeoutException while waiting for the state to update.

Modifications

• Modified OneWayReplicatorDeduplicationTest to set brokerDeduplicationSnapshotIntervalSeconds to 1 second in the test configuration.

• By increasing the frequency of the background snapshot monitor, the system is guaranteed to capture and persist the deduplication state frequently enough to satisfy the test's assertions, even if a specific manual trigger request is skipped due to a race condition.

Verifying this change

• Verified that the flaky test now runs consistently without failures.
• Confirmed that the 1-second background interval provides a sufficient "safety net" to persist the state when concurrent manual triggers are dropped.
• Make sure that the change passes the CI checks.

This change is already covered by existing tests, such as:

• org.apache.pulsar.broker.service.OneWayReplicatorDeduplicationTest.testDeduplication

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

[sandeep-mst]

LGTM
@Praveenkumar76 Assuming that the fix is tested well and is working as expected. As it's flaky, you should run it multiple times.

[Praveenkumar76]

yeah, I ran it multiple times locally, before raising and it looks stable