Configuration Lifecycle

From policy definition to fleet deployment—every step versioned, validated, and verifiable.

Hexarch manages configuration as immutable artifacts. You create, validate, bundle, deploy, and (if needed) roll back. The lifecycle is explicit, audited, and cryptographically verified.

The lifecycle

Define → Validate → Snapshot → Deploy → Monitor → (Rollback)

1. Define: Create or update policies, rules, and configurations
2. Validate: Artifact passes automated and manual checks
3. Snapshot: Bundle configuration into a named, immutable snapshot
4. Deploy: Push snapshot to target clusters
5. Monitor: Track fleet cohesion and drift
6. Rollback: Revert to previous snapshot if needed

Configuration artifacts

Artifacts are the building blocks—binary packages containing policy definitions:

Artifacts are immutable. You don’t edit artifacts—you create new versions.

Artifact validation

Before an artifact can be deployed, it must pass validation:

Automated validation (Hexarch-Scanner)

• Schema compliance
• Policy syntax verification
• Dependency resolution
• Security scanning

Manual audit (SecOps review)

• Business logic verification
• Compliance review
• Change impact assessment

Validation status is tracked:

• Validated: Passed all checks, deployable
• Pending: Awaiting validation
• Invalid: Failed checks, not deployable

Configuration snapshots

Snapshots bundle artifacts for deployment:

Snapshots are named and versioned. You deploy snapshots, not raw artifacts.

Snapshot metadata

Each snapshot includes:

• Change summary: Human-readable description of modifications
• Risk level: Low, Medium, High, Critical
• Approval requirement: Whether manual approval is required
• Delta impact: What changed from the previous snapshot

High-risk changes require explicit approval. The audit trail records who approved and when.

Fleet synchronization

When you deploy a snapshot:

1. Control plane broadcasts the desired state
2. Gateway nodes receive the snapshot
3. Nodes apply the configuration (hot-swap, no restart)
4. Nodes report their applied state
5. Control plane verifies hash match

If verification fails, the node is flagged as divergent. You see exactly which nodes failed and why.

Drift detection

Drift occurs when a node’s applied state doesn’t match the authority:

Causes:

• Failed hot-swap (e.g., incompatible filter signature)
• Network issues during sync
• Node resource pressure (memory, CPU)
• Manual tampering (rare but visible)

Detection:

• Continuous hash comparison
• Node-level status reporting
• Fleet cohesion metrics

Resolution:

• Automatic retry (if configured)
• Manual force sync with justification
• Node replacement (for persistent failures)

Rollback

If a deployment causes issues:

1. Select previous snapshot from history
2. Initiate rollback (requires justification)
3. Fleet syncs to the rollback snapshot
4. Verify cohesion across nodes

Rollbacks are deployments—they follow the same lifecycle and audit trail.

Snapshot states

Only one snapshot is Active per cluster at a time.

Why this matters

Without Configuration Lifecycle:

• “We deployed it” means hoping changes propagated
• Rollback is manual and error-prone
• Drift goes undetected
• Audit trails are incomplete

With Hexarch:

• Deployment is verified with cryptographic proof
• Rollback is one click (with audit)
• Drift is visible immediately
• Every change is recorded and attributable

Next steps

• Fleet Governance — how nodes are managed and verified
• Cryptographic Audit Chains — how deployments are recorded
• API Reference — endpoints for snapshot management