Skip to main content

Policies

The Policies screen (page heading Cryptographic policy) shows the rules every certificate must satisfy and lets you change them safely. Policies are signed and versioned, and both trust domains evaluate them.

ConsolePolicies

What it's for​

  • See the effective rules: allowed algorithms, maximum lifetimes, node quorum, deployment paths and issuance guardrails.
  • Configure the next policy version, including workload profiles.
  • Assess what the change would do to your managed estate before submitting it.
  • Activate a proposal. A different person must approve it.

How policies combine​

Three layers are intersected for every issuance:

  1. SECTIGO PLATFORM: an immutable safety floor controlled by Sectigo, outside your administration.
  2. ENTERPRISE: your tenant policy: algorithms, lifetimes, quorum and transition plan.
  3. WORKLOAD: the exact profile selected by the request: SPIFFE identity, DNS namespace, adapter, evidence and rotation rules.

Most restrictive policy wins. A lower layer can narrow authority but never expand it. A request is denied unless every layer allows it.

What you see​

Policies screen with the Configure next version button, the effective policy stack, the active policy details, effective constraints and the policy review queue
The Policies screen.

Effective policy stack​

Shows the three layers and the combined result.

FieldMeaning
Effective algorithmsSignature algorithms allowed by every layer. None — issuance blocked if no algorithm survives.
Maximum lifetimeSmallest lifetime ceiling, in hours.
Minimum customer quorumLargest required number of distinct node approvals.
Deployment pathsDeployment modes allowed by both the platform and the profile.
Issuance guardrailsEffective mode, requests per 5 minutes and SANs per day.

The status in the card header shows whether the Sectigo platform policy is in force:

StatusMeaning
independently enforcedThe platform policy is active and its digest is known.
platform policy unavailableThe platform policy could not be loaded. Issuance stays closed.
platform policy … (for example retired)The platform policy is not active.
platform policy expiredThe platform policy has passed its expiry.

Active policy​

The card title is the policy ID and version, for example acme-production · version 7. The status is active, draft, retired, expired or not yet effective.

BlockContent
PREFERRED SIGNATURESFirst-choice signature algorithms.
ALLOWED FALLBACKSOther allowed algorithms.
KEY ESTABLISHMENTPreferred key-exchange methods.
TRANSITIONTransition mode and post-quantum deadline.
ISSUANCE BEHAVIORGuardrail mode (or Inherited from the platform), requests per 5 minutes, SANs per 24 hours, and whether algorithm downgrade is blocked or monitored.
WORKLOAD / APPLICATION PROFILESEach profile with its maximum lifetime, and evidence if it requires capability evidence. No scoped profiles — issuance fails closed if there are none.

Effective constraints​

Maximum validity (days), Replica approval (distinct nodes), Offline renewal (bounded or disabled), Policy expires and Content binding (the start of the policy's SHA-256 digest; every component must agree on this exact digest).

Policy review queue​

Lists proposals with the number pending. Each proposal shows its version, policy ID, who proposed it and when, its digest, its status (pending, approved or rejected) and an impact badge:

  • Impact committed with x/y endpoints compatible, or
  • Compatibility blocked.

The action button on a pending proposal tells you what is possible:

ButtonMeaning
Approve & activateYou can activate this proposal.
Different reviewer requiredYou proposed it. Someone else must approve.
Resolve compatibility blockersThe impact assessment is blocking. It cannot be activated.
Refresh impact requiredThe proposal has no impact assessment. The proposer must refresh it.
Refresh estate impactShown to the proposer. Re-opens the exact proposal so you can re-assess the current estate.
Use as rollback baseShown on an earlier approved version. Opens the editor pre-filled from that version.

If there are no proposals: No change is awaiting review.

Configure the next version​

Configure policy version dialog with identity, lifetime and quorum fields, signature and key establishment choices, issuance guardrails and workload profiles
The policy editor.
  1. Select Configure next version. The editor opens as Configure policy version N, where N is the active version plus one. The tenant and version are fixed.
  2. Edit the sections you need (see the field reference below).
  3. Under Activation impact, select Assess managed estate.
  4. Review the verdict, the counts and any endpoints listed as not compatible.
  5. Select I reviewed this exact assessment….
  6. Select Submit committed impact.

The console confirms: Proposal submitted. A different authorized SSO user must approve activation.

Any edit invalidates the assessment

If you change the draft after assessing it, submission is rejected with The draft changed after assessment. Re-run estate impact before submission. Select Re-assess current draft and acknowledge again.

Identity, lifetimes, and quorum​

FieldRange
Policy IDText
ExpiresDate. Defaults to the later of the current expiry and one year from now.
Default lifetime (hours)1–8,760
Maximum lifetime (hours)1–8,760
Offline renewal ceiling (hours)0–8,760. 0 disables offline renewal.
Distinct node approvals1–5. Never lower than the Sectigo platform minimum.

Certificate signatures​

Set each algorithm to Preferred, Allowed fallback or Forbidden. Preferred is first choice, allowed is a fallback, and forbidden fails closed.

ECDSA-P256-SHA256, ECDSA-P384-SHA384, RSA-PSS-3072-SHA384, ML-DSA-65, SLH-DSA-SHA2-128s, HYBRID-ECDSA-P384-ML-DSA-65.

Key establishment and PQC transition​

FieldValues
X25519, X25519MLKEM768Preferred, Allowed fallback, Disabled
Transition modeClassical, Hybrid migration, PQC required
PQC required afterOptional date

Introduce hybrid transport before making it mandatory.

Issuance behavior guardrails​

Detect or block abnormal requester velocity, namespace expansion, algorithm downgrade and lifetime escalation. Values start from your current policy, or from the Sectigo platform defaults if you have none.

FieldValues
Response modeEnforce · block matching requests or Monitor · issue with signed finding
Requests per requester / 5 min1–1,000
Unique SANs per requester / 24 h1–10,000
Maximum lifetime increase (%)0–1,000
History lookback (days)1–90
Algorithm downgradeCheckbox: Signal a request weaker than the latest issued generation

Matching requests in enforce mode are denied before any CA is called. Findings appear on Incidents.

Workload and application profiles​

Each profile narrows who can request certificates and what they can get. Select Add profile to add one; it is named new-profile-1 (and so on) and starts from the scopes of your existing profiles. Remove is disabled when only one profile is left.

FieldMeaning
Profile IDThe profile name requests refer to. Must be unique.
Maximum lifetime (hours)Up to the policy's maximum lifetime.
Capability evidenceRequire fresh endpoint evidence before issuance.
SPIFFE path prefixesOne per line (or comma-separated), for example spiffe://your-trust-domain/workloads/.
DNS suffixesOne per line (or comma-separated). Converted to lower case.
AlgorithmsWhich of the policy's allowed algorithms this profile may use.
Deployment pathssingle, dual slot, microsoft adcs, nginx, apache, java keystore, hashicorp vault. Choosing a dual-slot-capable path switches the profile to dual-slot rotation.
PromotionOperator verifies & promotes, or Automatic after exact health (only with NGINX, Apache, Java keystore or HashiCorp Vault).
Renew before (hours)From 1 up to the profile's maximum lifetime.
Overlap (minutes)0–10,080
Health grace (seconds)10–3,600
Rollback window (minutes)1–1,440

Eligible approval nodes​

Leave all nodes unchecked to let any enrolled node count toward the quorum, or check specific nodes to pin an explicit quorum set.

Activation impact​

Activation impact section with the assessment verdict, compatibility counts, commitments and the acknowledgement checkbox
An estate impact assessment ready for review.

The quick summary, computed before you assess:

LineMeaning
Algorithms removedAllowed algorithms this version drops, or None.
Healthy-node quorum nowx available / y required. Highlighted if too few nodes are healthy. You cannot submit until enough nodes are healthy.
Maximum certificate lifetimeIn hours.
Offline renewalHours, or Disabled.
Workload profilesNumber of profiles.
Dependency pathsLive and committed edges after assessment.

Assess managed estate computes a server-side snapshot from your certificate sets, in-flight issuance and latest capability evidence. The verdict reads Assessment is ready for independent review or Cryptographic activation is blocked, followed by whether cryptographic behavior changes in this version.

Each managed endpoint is classified:

ClassMeaning
compatibleFresh evidence shows the endpoint supports what the platform, the proposed policy and exactly one profile allow.
incompatibleNo allowed signature or key-exchange overlap, or hybrid/PQC support is not measured.
missingThe endpoint has no capability record.
staleIts newest evidence has expired.
unassignedNo profile owns the endpoint's identity.
ambiguousMore than one profile owns it.
unverifiedIts newest evidence was imported by an administrator, not observed by a machine.

A change to signatures, key establishment, transition mode or workload profiles is blocked while any endpoint is not compatible, including when the estate is empty. Up to eight problem endpoints are listed with reasons; large estates note that details are bounded but the commitment covers every endpoint.

The assessment ends with four digests: Platform baseline, Estate commitment, Dependency graph and Impact commitment. If Platform baseline reads Not bound — refresh required, re-assess.

Finally, select the acknowledgement: I reviewed this exact assessment. Any capability refresh, expiry, endpoint, profile, or evidence change will invalidate approval.

Advanced JSON preview shows the full draft policy.

Approve and activate a proposal​

  1. Sign in as a different person from the proposer.
  2. Open ConsolePolicies and find the proposal in Policy review queue.
  3. Check the impact badge reads Impact committed.
  4. Select Approve & activate.

The console confirms: Policy version N activated and queued for signed node distribution. The Policy version tile on Trust health updates.

Activation is rejected if anything changed since the assessment: a new endpoint, new or expired evidence, a dependency change, a profile mapping or the Sectigo platform baseline. The proposer must then select Refresh estate impact, re-assess and acknowledge again.

Roll back to an earlier version​

There is no "revert" switch: a rollback is a new version that a second person activates, like any other change.

  1. In Policy review queue, find the earlier approved version and select Use as rollback base. The editor opens pre-filled with that version's content.
  2. Expand Advanced JSON preview and note the settings you want to restore. Then close the editor.
  3. Select Configure next version and re-enter those settings.
  4. Assess, acknowledge and submit as usual. A second person activates it.
Why not submit the rollback draft directly?

In the current release, a draft opened with Use as rollback base is numbered from the older version rather than from the active one, so assessing it fails with Policy must preserve the tenant and advance exactly one version. Use it as a reference and submit the change from Configure next version.

Platform policy limits​

The Sectigo platform baseline cannot be changed from the editor. The editor shows it as platform-id@version. Your policy can make it stricter but never weaker:

  • You cannot allow an algorithm the platform forbids.
  • Lifetimes and the offline renewal window cannot exceed the platform ceiling.
  • Distinct node approvals cannot be lower than the platform minimum.
  • Deployment paths must be allowed by the platform.
  • Guardrails can only be tightened: enforce mode, lower ceilings, downgrade protection and longer lookback are stricter.

A draft that leaves no allowed signature, key-establishment method or deployment path is rejected before assessment.

Permissions​

ActionRoles
View the active policy and stackEvery role
View the review queue, assess impactPKI Operator, Tenant Admin
Submit a proposalPKI Operator, Tenant Admin. Requires MFA within the last ten minutes.
Approve & activatePKI Operator, Tenant Admin, and a different person from the proposer. Requires MFA within the last ten minutes.

See Policy changes & approvals for the operational process.

Troubleshooting​

Message or symptomCause and fix
Assess and acknowledge the exact estate impact before submissionRun the assessment and select the acknowledgement.
The draft changed after assessment. Re-run estate impact before submission.Select Re-assess current draft, then acknowledge again.
Policy must preserve the tenant and advance exactly one versionThe draft is not the next version of the active policy. Close the editor and select Configure next version again.
Workload profile … already existsChoose a unique Profile ID.
Submit committed impact is greyed outThe assessment is not acknowledged, or fewer nodes are healthy than Distinct node approvals requires.
Policy proposals could not be loaded or a permission message above the stackYour role cannot manage policy. You can still view the active policy.
Different reviewer requiredYou proposed this version. Ask another PKI Operator or Tenant Admin.
Resolve compatibility blockersFix the endpoints listed in the assessment (refresh evidence, assign profiles) or change the draft, then submit a new proposal.