State Machine Replication: Log Design, Snapshots, and Compaction

Monthly research note. Theme: Distributed Systems Under Failure.

TL;DR

A focused memo on State Machine Replication: Log Design, Snapshots, and Compaction: define the model, state the properties, then design the system so those properties remain true under failure and adversaries.

Key takeaways

• Expose protocol state (epoch/term/commit index) as first-class telemetry.
• Treat membership changes and compaction as protocol events—not operational details.
• Backpressure and admission control are correctness mechanisms under load.
• Treat retries, reordering, and partial failure as default conditions.
• Bind security decisions to evidence (audit, invariants, telemetry).

Why this matters

• Safety failures are permanent; liveness failures are (sometimes) recoverable.
• Global systems fail in correlated ways (regions, dependencies, routing).
• If your protocol isn’t testable under reordering, it isn’t deployable.
• Observability must explain protocol state, not just latency.

Key questions

• What does “read” mean under replication lag?
• What is the unit of ordering (per key, per partition, global)?
• What is your reconfiguration model (joint consensus, epochs, leases)?
• How do clients discover leaders safely (and what happens during flaps)?
• Where do you pay for liveness (timeouts, leader election, reconfiguration)?
• How do you prevent overload from becoming inconsistency?

Assumptions

• Workload is skewed: hot keys exist and dominate.
• Nodes restart with partial state unless you prove durability.
• Clients retry and amplify load right when the system is weakest.
• Packets can be duplicated and reordered; acks can be lost.

Non-goals

• Pretending backpressure is an implementation detail.
• Assuming the network eventually behaves “nicely” under load.

Model & invariants

A common safety shape for replicated logs:

Liveness is always conditional: specify when progress is expected and what you do otherwise.

Treat membership changes as protocol events, not control-plane side effects.

Security properties

• Downgrade resistance: negotiation can’t silently weaken security posture.
• Evidence: critical actions emit verifiable audit events.
• Authenticity: actions are bound to identity and purpose.
• Replay resistance: duplicated inputs do not change outcomes.

Failure modes

• Mixed-version behavior that violates assumptions silently.
• Observability gaps during incidents (missing evidence).
• Resource exhaustion (CPU/bandwidth/storage) turning into correctness failures.
• Timeout ambiguity causing double-apply or partial state transitions.

Design sketch

flowchart TD
  client["Client"] --> leader["Leader"]
  leader --> log["Replicated Log"]
  log --> snap["Snapshot"]
  snap --> recover["Recovery / Catch-up"]
  leader --> reconfig["Reconfiguration"]

Implementation notes

Your protocol is an interface between failures and invariants. Encode both.

Operational invariants to monitor:
- leader_changes_per_minute
- commit_index_monotonic
- snapshot_install_failures
- quorum_acks_latency_p99
- rejected_requests_due_to_admission_control

Verification strategy

• Jepsen-style fault injection: partitions + reordering + client retries.
• Upgrade tests: mixed versions and rolling deploy invariants.
• Deterministic replay of network traces to reproduce rare failures.
• Model checking the smallest core (timeouts, election, reconfiguration).
• Stress + skew tests: hot keys, slow disks, noisy neighbors.

Operational notes

• Expose protocol state: term/epoch, leader, commit index, config version.
• Make client behavior part of the system: document retry semantics.
• Treat compaction and snapshot install as first-class SLOs.
• Rehearse region failover and reconfiguration under load.
• Rate-limit retries and apply admission control before saturation.

What to monitor

• Rollback events and the conditions that triggered them.
• Error budget burn + tail latency under load.
• Authz failures and policy denials (unexpected spikes).
• Admission-control / rate-limit rejections (by reason).
• Invariant violation rate (should be ~0).

Rollback plan

• Keep dual-write / dual-verify windows where appropriate.
• Prefer backward-compatible changes; avoid “flag day” upgrades.
• Use canaries and staged rollout; stop early when signals degrade.
• Define an explicit rollback trigger (metrics + thresholds).
• Preserve evidence (configs, artifacts, audit logs) to reconstruct what changed.

Evidence

• Time, Clocks, and the Ordering of Events (Lamport) (1) — Causality, ordering, and why clocks are tricky.
  • Evidence: Use this as the baseline for happens-before vs wall-clock; avoid embedding clock assumptions into safety properties.
• Jepsen (2) — Testing correctness under partitions and faults.
  • Evidence: Turn faults into test cases; prioritize partition and clock-skew scenarios that violate user-visible guarantees.

Open questions

• Which invariants are violated first under overload: latency, availability, or correctness?
• What is the worst-case recovery time after a leader + disk failure?
• How do you prevent “operator fixes” from changing safety properties?
• Where does your protocol assume synchrony without admitting it?

Checklist

• Telemetry captures correctness signals.
• Rollback plan rehearsed and automated.
• Assumptions listed and reviewed.
• Safety properties stated as invariants.
• Costs bounded (CPU/memory/bandwidth) under adversarial inputs.
• Failure modes enumerated with mitigations.

Further reading

• Jepsen — Testing correctness under partitions and faults.
• Time, Clocks, and the Ordering of Events (Lamport) — Causality, ordering, and why clocks are tricky.
• Paxos Made Simple (Lamport) — Agreement basics and the invariants that matter.
• In Search of an Understandable Consensus Algorithm (Raft) — Consensus with explicit state machines and practical tradeoffs.
• Learn TLA+ — Practical entry point for specification and model checking.
• Site Reliability Engineering (Google) — Error budgets, incident response, and reliability as an engineering discipline.