Time, Clocks, and the Ordering of Events in a Distributed System.

Lamport, L. (1978). Time, Clocks, and the Ordering of Events in a Distributed System. Communications of the ACM, 21(7), 558-565.

Type

Journal article

Intellectual base

Primary research

Year

1978

DOI

10.1145/359545.359563

Link

https://doi.org/10.1145/359545.359563

Cited from

computer_science, music_musicology, neuroscience, physics, political_science, systems_cybernetics

Cited by

9 citations across 8 artifacts.

Each citation links to the sentence it supports in the citing article.

Primes

• Concurrency
  • Time-ordering sensitivity and causality preservation across parallel execution

    This sourceDefines the happened-before partial order and logical clocks for ordering distributed events. SUPPORTS marker 049 (time-ordering sensitivity and causality preservation). PARTIALLY SUPPORTS marker 059 (T3: determinism vs nondeterminism) — Lamport gives the causal-ordering framework but does not directly treat testing-interleaving nondeterminism; defensible but indirect. DOI verified. See flag on 059.

• Inconsistent Shared Model
  • Introducing a reconciliation protocol the subsystems run voluntarily and recurrently — two-phase commit, vector clocks, consensus protocols, fact-check meetings, sensor-fusion arbiters with explicit uncertainty, CRDTs — is the same move again.

    This sourceLogical clocks (vector-clock ancestor) for reconciling event orderings across uncoupled processes — a reconciliation protocol run recurrently between subsystems.

• Local Autonomy & Tiered Escalation
  • When well-designed, tiered systems are fast (Tier-1 resolves most issues immediately), efficient (specialists focus on hard problems), and adaptive (escalation criteria can be tuned as conditions change), much as Lamport (1978) demonstrates in his analysis of clocks and event ordering in distributed systems.

    This sourceFoundational analysis of distributed systems: local processes act autonomously on local clock state and coordinate only when causal ordering across processes requires it — the distributed-computing analogue of tiered local autonomy with selective escalation.

• Polyphony
  • Deliberation and pluralist politics. Multi-stakeholder fora and citizen assemblies in which distinct positions must remain audible as their own arguments rather than being pre-aggregated into a single majority view. Organizations. Cross-functional groups in which engineering, design, legal, and sales each speak in their own register; the shared "score" is the product or strategy, not a single hierarchical voice. Distributed systems. Independent processes or services running concurrently on a shared bus or coordination layer, where correctness depends on each holding its own state machine while honoring shared invariants.

    This sourceFoundational account of independent concurrent processes coordinating on a shared logical-time substrate while each holds its own state.

• Race Condition
  • A second abstract move is that races are sensitive to substrate-level guarantees that are invisible at the level of intent: the same protocol can be race-free on a synchronous substrate and race-prone on an asynchronous one, which is the structural reason that moving a working system to a more loosely coordinated environment can introduce races silently in logic that never changed.

    This sourceShows that ordering of concurrent events depends on substrate-level guarantees and that bounded delay / logical clocks are needed even to reason about order.

• Synchronization
  • Computer science & distributed systems: Network Time Protocol (NTP) and Precision Time Protocol (PTP) synchronize clocks across geographically dispersed servers; Lamport (1978) logical clocks and vector clocks provide causality without requiring physical time agreement; consensus protocols (Paxos, Raft) solve the distributed-consensus problem by achieving agreement on state despite asynchrony; thread synchronization (locks, barriers, semaphores, condition variables) ensures that concurrent threads access shared resources in consistent order.

    This sourceFoundational analysis of distributed systems: local processes act autonomously on local clock state and coordinate only when causal ordering across processes requires it — the distributed-computing analogue of tiered local autonomy with selective escalation.

  • But modern consensus protocols (Paxos, Raft) go further: they achieve agreement on state order using logical clocks, without requiring tight physical synchronization—an architecture made possible by Lamport's (1978) demonstration that the happens-before relation is sufficient for distributed causality.

    This sourceDemonstrates that consensus and state-order agreement in distributed systems can be achieved using logical clocks alone, without tight physical clock synchronization—the theoretical foundation underlying Paxos and Raft.

• Temporal Synchronization and Phase Alignment
  • In the first case, synchronization suffices as a concept; in the second case, phase alignment is the critical structural property.

    This sourceFoundational analysis of distributed systems: local processes act autonomously on local clock state and coordinate only when causal ordering across processes requires it — the distributed-computing analogue of tiered local autonomy with selective escalation.

Domain-specific

• Happened-before
  • Lamport's happened-before relation a→b holds when a precedes b in one process, a sends a message received at b, or a→c→b; unrelated events are concurrent.

Verification

This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.

Registry ID ref:568bb47d565c · see in the full table