Oftentimes, database systems are made up of processes (also known as threads) that communicate with one another via interacting with shared memory resources. Access to the shared memory space is restricted in what way, if at all. When it comes to serializing access to data structures in shared memory, is two-phase locking the most acceptable technique? Your response should be supported by reasoning.

How does "priority inversion" impact database concurrency control mechanisms?

Distributed database management systems face the same concurrency control challenges as other categories of distributed systems. Is it necessary to employ concurrency control mechanisms such as two-phase locking (also known as 2PL) in a configuration with no duplicates? Explain.

How do databases manage the "phantom read" problem in concurrency control?

Take, for example, a database that is used by a large airline and in which the underlying database management system utilises snapshot isolation. In spite of the fact that the airline would prefer a serializable execution, it could be ready to tolerate a non serializable execution if it means that the process would be completed more quickly overall.

The same concurrency management issues apply to distributed database systems as they do to other distributed systems. Is it necessary to use 2PL (Two-phase locking) concurrency control techniques for a system that does not have any duplicated components? Explain.

Assume that a database system is installed on a server with 3 core processors. You want your query to run on multiple processors so that each processor accesses in parallel a subset of the database. Or, you might set up your system in a way that partitions your query so that each part of the same query runs on a different processor. Explain, with an example, how a query might be broken apart so that each parallel processor handles a subset of the query

Concurrency management is a challenge in distributed database systems, as it is in any distributed system. Is 2PL (two-phase locking) necessary if there are no duplicated fragments? Explain.

This inquiry pertains to the distinction between database replication and database fragmentation.

How do "priority-based" concurrency control mechanisms ensure fair access to database resources?