Principles Of Distributed Database Systems Exercise Solutions [portable] Direct

Each site manages its own locks. T1 locks A at site1. T2 locks B at site2. T1 sends lock request for B to site2 → waits. T2 sends lock request for A to site1 → waits. Deadlock is distributed. Needs timeout or probe-based detection (e.g., wait-for graph across sites).

For example, if a new employee is added at Site A, the employee's information is stored in the local database at Site A. If the employee's department is updated at Site B, the updated information is stored in the local database at Site B. The system ensures that the data is consistent across all sites by using distributed transactions and concurrency control. Each site manages its own locks

Initial inventory = 5 at all replicas.

Transparency ensures the user perceives the system as a single, centralized database. T1 sends lock request for B to site2 → waits