Environment
- Infinispan 9.4.18 embedded
- 3 nodes cache in replicated mode
- RocksDB store (or other, doesn't matter)
Steps to reproduce
- Create TCP-based cluster
- Create cache
- Add Entity to cache
- Check that entity is stored on each node
- Stop non-coordinator node
- Remove Entity from cache on coordinator
- Check that Entity is removed on running nodes
- Start previously stopped non-coordinator node
- Check that Entity is present on restarted node but is absent on remaining nodes. But expected behavior is to replicate removing to restarted node too.
Questions 1. Is this behavior OK? 2. Can I change it to indicated as expected? 3. And how if so?