Subtopic Deep Dive
Concurrency Control in Distributed Databases
Research Guide
What is Concurrency Control in Distributed Databases?
Concurrency control in distributed databases ensures data consistency under concurrent transactions across geo-replicated nodes using protocols like two-phase locking, optimistic concurrency control, and multi-version concurrency control.
This subtopic addresses serializability guarantees, performance trade-offs, and integration with replication in systems like Spanner. Key works include Bernstein et al. (1987) with 4310 citations on concurrency mechanisms and Herlihy and Wing (1990) with 3150 citations defining linearizability for concurrent objects. Over 10 highly cited papers from 1985-2000 establish foundational protocols.
Why It Matters
Concurrency control enables scalable data management in global systems like CockroachDB by preventing conflicts in geo-replicated environments. Bernstein et al. (1987) detail locking and recovery for transaction processing, applied in enterprise databases. Herlihy and Wing (1990) provide linearizability for high-concurrency shared objects, influencing modern NoSQL stores. Lamport (1998) supports consistent replication via Paxos, critical for fault-tolerant banking and e-commerce platforms.
Key Research Challenges
Achieving Serializability
Distributed transactions require serializability without centralized coordination, leading to aborts or latency. Bernstein et al. (1987) analyze two-phase locking limitations in networks. Optimistic methods increase retries under contention.
Balancing Latency and Consistency
Geo-replication demands low latency with strong guarantees like linearizability. Herlihy and Wing (1990) define linearizability but note performance costs. Protocols like Paxos from Lamport (1998) add overhead for fault tolerance.
Fault Tolerance Integration
Crashes and partitions complicate locking and validation. Fischer et al. (1985) prove consensus impossibility with faults, impacting control protocols. Chandra and Toueg (1996) introduce failure detectors to enable solutions.
Essential Papers
Impossibility of distributed consensus with one faulty process
Michael J. Fischer, Nancy Lynch, Michael S. Paterson · 1985 · Journal of the ACM · 4.5K citations
The consensus problem involves an asynchronous system of processes, some of which may be unreliable. The problem is for the reliable processes to agree on a binary value. In this paper, it is shown...
Concurrency Control and Recovery in Database Systems
Philip A. Bernstein, Vassco Hadzilacos, Nathan Goodman · 1987 · 4.3K citations
This book is an introduction to the design and implementation of concurrency control and recovery mechanisms for transaction management in centralized and distributed database systems. Concurrency ...
Hyperledger fabric
Elli Androulaki, Artem Barger, Vita Bortnikov et al. · 2018 · 3.2K citations
Fabric is a modular and extensible open-source system for deploying and operating permissioned blockchains and one of the Hyperledger projects hosted by the Linux Foundation (www.hyperledger.org). ...
Linearizability: a correctness condition for concurrent objects
Maurice Herlihy, Jeannette M. Wing · 1990 · ACM Transactions on Programming Languages and Systems · 3.1K citations
A concurrent object is a data object shared by concurrent processes. Linearizability is a correctness condition for concurrent objects that exploits the semantics of abstract data types. It permits...
System architecture directions for networked sensors
Jason Hill, Robert Szewczyk, Alec Woo et al. · 2000 · ACM SIGPLAN Notices · 3.1K citations
Technological progress in integrated, low-power, CMOS communication devices and sensors makes a rich design space of networked sensors viable. They can be deeply embedded in the physical world and ...
The part-time parliament
Leslie Lamport · 1998 · ACM Transactions on Computer Systems · 2.7K citations
Recent archaeological discoveries on the island of Paxos reveal that the parliament functioned despite the peripatetic propensity of its part-time legislators. The legislators maintained consistent...
Unreliable failure detectors for reliable distributed systems
Tushar Chandra, Sam Toueg · 1996 · Journal of the ACM · 2.5K citations
We introduce the concept of unreliable failure detectors and study how they can be used to solve Consensus in asynchronous systems with crash failures. We characterise unreliable failure detectors ...
Reading Guide
Foundational Papers
Start with Bernstein et al. (1987) for comprehensive CC mechanisms in distributed systems, then Fischer et al. (1985) for consensus impossibilities impacting protocols, followed by Herlihy and Wing (1990) for linearizability definitions.
Recent Advances
Lamport (1998) on Paxos for practical replication; Chandra and Toueg (1996) on failure detectors enabling reliable CC.
Core Methods
Two-phase locking, optimistic concurrency control with validation, multi-version CC, linearizability checks, integrated with Paxos consensus.
How PapersFlow Helps You Research Concurrency Control in Distributed Databases
Discover & Search
Research Agent uses searchPapers for 'concurrency control distributed databases' to find Bernstein et al. (1987), then citationGraph reveals 4310 citing works and back-references to Fischer et al. (1985). exaSearch uncovers geo-replication extensions; findSimilarPapers links Lamport (1998) to modern Spanner papers.
Analyze & Verify
Analysis Agent applies readPaperContent to Bernstein et al. (1987) for protocol details, verifyResponse with CoVe checks serializability claims against Herlihy and Wing (1990), and runPythonAnalysis simulates locking contention with NumPy for throughput stats. GRADE scores evidence strength on linearizability proofs.
Synthesize & Write
Synthesis Agent detects gaps in optimistic control for geo-replication, flags contradictions between Lamport (1998) and failure models. Writing Agent uses latexEditText for protocol pseudocode, latexSyncCitations for 10 foundational papers, latexCompile for report, and exportMermaid diagrams locking states.
Use Cases
"Simulate two-phase locking performance in high-contention distributed DB"
Research Agent → searchPapers 'two-phase locking distributed' → Analysis Agent → runPythonAnalysis (pandas model of 1000 txns, plot abort rates) → matplotlib throughput graph.
"Write LaTeX section comparing linearizability vs serializability in geo-DBs"
Research Agent → citationGraph on Herlihy and Wing (1990) → Synthesis → gap detection → Writing Agent → latexEditText draft → latexSyncCitations (Bernstein 1987) → latexCompile PDF.
"Find GitHub code for Paxos-based concurrency control"
Research Agent → searchPapers 'Paxos concurrency control' (Lamport 1998) → Code Discovery → paperExtractUrls → paperFindGithubRepo → githubRepoInspect (test suite, benchmarks).
Automated Workflows
Deep Research workflow scans 50+ papers via searchPapers on 'distributed concurrency control', structures report with GRADE-verified sections on protocols from Bernstein et al. (1987). DeepScan applies 7-step analysis to Lamport (1998), checkpointing CoVe on consensus claims. Theorizer generates hypotheses on failure detectors from Chandra and Toueg (1996) integrated with optimistic CC.
Frequently Asked Questions
What defines concurrency control in distributed databases?
Mechanisms like two-phase locking, optimistic validation, and multi-version CC ensure serializability across nodes (Bernstein et al., 1987). They manage conflicts in geo-replicated transactions.
What are core methods?
Two-phase locking avoids deadlocks; optimistic CC defers checks; linearizability provides real-time guarantees (Herlihy and Wing, 1990). Paxos aids replication consistency (Lamport, 1998).
What are key papers?
Bernstein et al. (1987, 4310 citations) covers mechanisms; Fischer et al. (1985, 4508 citations) proves consensus limits; Herlihy and Wing (1990, 3150 citations) defines linearizability.
What open problems exist?
Reducing latency in strongly consistent geo-replication under faults. Integrating failure detectors (Chandra and Toueg, 1996) with low-overhead CC remains challenging.
Research Distributed systems and fault tolerance with AI
PapersFlow provides specialized AI tools for Computer Science researchers. Here are the most relevant for this topic:
AI Literature Review
Automate paper discovery and synthesis across 474M+ papers
Code & Data Discovery
Find datasets, code repositories, and computational tools
Deep Research Reports
Multi-source evidence synthesis with counter-evidence
AI Academic Writing
Write research papers with AI assistance and LaTeX support
See how researchers in Computer Science & AI use PapersFlow
Field-specific workflows, example queries, and use cases.
Start Researching Concurrency Control in Distributed Databases with AI
Search 474M+ papers, run AI-powered literature reviews, and write with integrated citations — all in one workspace.
See how PapersFlow works for Computer Science researchers