Summary

This PR fixes two critical bugs that were causing e2e test failures due to concurrent operation conflicts:

1. Policy Event Handler - Duplicate Events Issue

Problem: PostgreSQL error "ON CONFLICT DO UPDATE command cannot affect row a second time" occurred when duplicate events with the same composite key (event_name, count, created_at) were present in a single batch insert

Solution: Added deduplication logic in policy event handlers using a map to filter duplicate events before batch insertion, keeping only the last occurrence of duplicates

2. KafkaUser Update - Concurrent Modification Issue

Problem: Massive KafkaUser update conflicts with error "the object has been modified; please apply your changes to the latest version and try again". In 2 seconds (12:10:00-12:10:02), operator attempted to update hub1-kafka-user over 20 times, all failing due to resource version conflicts

Solution: Added retry logic with retry.RetryOnConflict to the EnsureUser method to handle optimistic concurrency conflicts when multiple controllers try to update the same KafkaUser resource simultaneously

Changes

• Deduplicate events in local_root_policy_event_handler.go and local_replicated_policy_event_handler.go before batch insertion
• Add retry mechanism in KafkaUser update operations to handle concurrent modifications gracefully

PR Link

https://github.com/stolostron/multicluster-global-hub/pull/2044