Abstract:A DBMS allows trading consistency for efficiency through the allocation of isolation levels that are strictly weaker than serializability. The robustness problem asks whether, for a given set of transactions and a given allocation of isolation levels, every possible interleaved execution of those transactions that is allowed under the provided allocation, is always safe. In the literature, safe is interpreted as conflict-serializable (to which we refer here as conflict-robustness). In this paper, we study the view-robustness problem, interpreting safe as view-serializable. View-serializability is a more permissive notion that allows for a greater number of schedules to be serializable and aligns more closely with the intuitive understanding of what it means for a database to be consistent. However, view-serializability is more complex to analyze (e.g., conflict-serializability can be decided in polynomial time whereas deciding view-serializability is NP-complete). While conflict-robustness implies view-robustness, the converse does not hold in general. In this paper, we provide a sufficient condition for isolation levels guaranteeing that conflict- and view-robustness coincide and show that this condition is satisfied by the isolation levels occurring in Postgres and Oracle: read committed (RC), snapshot isolation (SI) and serializable snapshot isolation (SSI). It hence follows that for these systems, widening from conflict- to view-serializability does not allow for more sets of transactions to become robust. Interestingly, the complexity of deciding serializability within these isolation levels is still quite different. Indeed, deciding conflict-serializability for schedules allowed under RC and SI remains in polynomial time, while we show that deciding view-serializability within these isolation levels remains NP-complete.

What problem does this paper attempt to address?

### What problem does this paper attempt to solve? This paper mainly explores the relationship between view - robustness and conflict - robustness in Multiversion Concurrency Control (MVCC). Specifically, the authors study the problem of whether these two types of robustness are consistent under a given isolation level. #### Main problems 1. **The equivalence of view - robustness and conflict - robustness**: - Conflict - robustness means that under a given isolation level, all possible interleaved executions of transactions are conflict - serializable. - View - robustness means that under a given isolation level, all possible interleaved executions of transactions are view - serializable. - The core question of the paper is: for some isolation levels, are view - robustness and conflict - robustness always equivalent? If they are equivalent, then under these isolation levels, expanding from conflict - serializable to view - serializable will not allow more sets of transactions to become robust. 2. **Complexity analysis**: - Deciding conflict - serializability can be done in polynomial time, while deciding view - serializability is an NP - complete problem. - The paper also explores the complexity differences between these two types of robustness under the isolation levels (read - committed, snapshot isolation, serializable snapshot isolation) in PostgreSQL and Oracle. #### Research background - In concurrency control, serialization is the most ideal isolation level, but it is usually very costly. - In practical applications, database systems offer multiple weaker isolation levels in exchange for higher performance. - These weaker isolation levels may lead to specific exceptions, but in some cases, a set of transactions can be safely executed under a weaker isolation level without causing errors. #### Main contributions 1. **Proposing a sufficient condition**: Ensure that under some isolation levels, view - robustness and conflict - robustness are consistent. 2. **Proving the isolation levels in PostgreSQL and Oracle** (read - committed, snapshot isolation, serializable snapshot isolation) satisfy this condition. 3. **Revealing the complexity differences**: Although view - robustness and conflict - robustness are consistent under these isolation levels, deciding view - serializability is still an NP - complete problem, while deciding conflict - serializability can be solved in polynomial time. Through these studies, the paper provides a new perspective for understanding robustness in concurrency control and a theoretical basis for further optimizing the concurrency control mechanism of database systems.

When View- and Conflict-Robustness Coincide for Multiversion Concurrency Control

Robustness against Read Committed for Transaction Templates with Functional Constraints

Rethinking serializable multiversion concurrency control

Multiversion Conflict Notion for Transactional Memory Systems

Automated Detection of Serializability Violations under Weak Consistency

On the complexity of checking transactional consistency

Repairing Conflicts among MVCC Transactions

Pushing the Limit: Verified Performance-Optimal Causally-Consistent Database Transactions

Decentralizing MVCC by Leveraging Visibility

SQL Access Patterns for Optimistic Concurrency Control

Improving High Contention OLTP Performance via Transaction Scheduling

Parameterized and Runtime-Tunable Snapshot Isolation in Distributed Transactional Key-Value Stores

A Critique of Snapshot Isolation

Batch-Schedule-Execute: On Optimizing Concurrent Deterministic Scheduling for Blockchains (Extended Version)

On The Design Of Adaptive And Speculative Concurrency Control In Distributed Databases

Multi-Valued Connected Consensus: A New Perspective on Crusader Agreement and Adopt-Commit

Read Consistency in Distributed Database Based on DMVCC

Topics in Cosmology

Coo: Consistency Check for Transactional Databases

Concurrency control in database systems

A TimeStamp Based Multi-version STM Algorithm