Reproducible CI/CD: Determinism as Defense

Monthly research note. Theme: DevSecOps & Resilience Engineering.

TL;DR

Reproducible CI/CD: Determinism as Defense as an engineering constraint: write down assumptions, make invariants executable, and design operational recovery as part of correctness.

Key takeaways

• Make rollback a first-class operation with explicit triggers and rehearsal.
• Policy-as-code needs tests, rollout, and rollback like any other production system.
• Treat CI/CD as attacker-controlled until proven otherwise; minimize secrets and privileges.
• Measure correctness signals, not only latency/throughput.
• Make boundaries boring: validate inputs, cap costs, and be deterministic where needed.

Why this matters

• Infrastructure-as-code without policy is just scripting the attack surface.
• Policy drift is the default; guardrails must be automated and enforced.
• Reproducibility is how you know what you shipped is what you built.
• Secrets in CI turn “one compromised job” into “full compromise.”

Key questions

• Which signals prove correctness (not just availability) in production?
• How do you rehearse incident response as code (runbooks, chaos, drills)?
• What is the minimum set of humans who can ship to production?
• What is your supply-chain threat model (dependency poisoning, CI compromise)?
• How do you manage secrets without long-lived credentials in CI?
• Where do you enforce policy (pre-merge, build, deploy, runtime)?

Assumptions

• CI runners are exposed to untrusted code (PRs, dependencies).
• Rollbacks must be executed under time pressure.
• Dependencies can be compromised upstream (typosquatting, maintainer takeover).
• Policy enforcement must be consistent across environments.

Non-goals

• Long-lived credentials embedded in pipelines.
• Manual policy enforcement or manual security review as the only control.

Model & invariants

Build provenance is a cryptographic statement:

Treat CI as attacker-controlled until proven otherwise; minimize secrets and privileges.

Make provenance verifiable: “what built this” must be cryptographically bound.

Security properties

• Evidence: critical actions emit verifiable audit events.
• Least authority: privileges are scoped by purpose and time.
• Integrity: invalid transitions are rejected (and detectable).
• Authenticity: actions are bound to identity and purpose.

Failure modes

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

Design sketch

flowchart LR
  src["Source"] --> build["Build (reproducible)"]
  build --> attest["Attestation"]
  attest --> scan["SAST/DAST/SCA"]
  scan --> deploy["Deploy (policy gates)"]
  deploy --> runtime["Runtime Policy + Observability"]

Implementation notes

The pipeline is production: it has credentials, network reach, and authority.

// Treat CI as untrusted: keep tokens short-lived and scoped.
type Token struct {
  Value string
  ExpiresAtUnix int64
  Scope string
}

Verification strategy

• Pipeline attack simulations: compromise a runner and measure blast radius.
• Rollback tests as part of release (not “if needed”).
• Dependency tampering drills: lockfile changes, integrity failures.
• Policy tests: unit tests for policy-as-code rules.
• Runtime conformance: detect drift between desired and actual state.

Operational notes

• Keep a provenance trail for every artifact deployed to production.
• Treat policy changes as security-sensitive deploys (review + rollout).
• Audit who can ship and how; remove implicit paths.
• Rehearse incident response for the pipeline itself.
• Continuously scan and inventory dependencies; prioritize by exposure.

What to monitor

• Invariant violation rate (should be ~0).
• Authz failures and policy denials (unexpected spikes).
• Error budget burn + tail latency under load.
• Rollback events and the conditions that triggered them.
• Retry/timeout rates by endpoint and client cohort.

Rollback plan

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

Evidence

• Designing Data-Intensive Applications (Kleppmann) (1) — The systems-engineering baseline for correctness, replication, and failure.
  • Evidence: Replication and consistency tradeoffs as engineering constraints; use as reference when naming guarantees.
• Learn TLA+ (2) — Practical entry point for specification and model checking.
  • Evidence: Model the smallest thing that can break; use model checking to validate invariants before optimizing.

Open questions

• Can you answer “what code is running” with cryptographic evidence?
• What is the smallest CI compromise that becomes a prod compromise today?
• How quickly can you revoke all pipeline credentials in an incident?
• Which deploy actions are irreversible and how do you mitigate that?

Checklist

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

Further reading

• SLSA v1.0 Specification — Supply-chain levels and provenance requirements.
• Sigstore — Signing and verifying artifacts at scale with transparency logs.
• NIST SP 800-218 (SSDF) — Secure software development practices as an engineering framework.
• in-toto — Securing the integrity of software supply chains with attestations.
• Designing Data-Intensive Applications (Kleppmann) — The systems-engineering baseline for correctness, replication, and failure.
• Learn TLA+ — Practical entry point for specification and model checking.