Operations and system health

Health, kill switches, wallets, audit search, tenant notices and support sessions.

System health#

Operations › System health shows dependency checks and recent provider observations: response times, waiting and failed jobs, free disk, running sends, delivery-report activity, operator queue depths, and each provider's circuit state, delivery rate, latency and balance alert. Choose Refresh now for a new sample.

A zero queue does not prove a worker is running. Use the production monitoring alerts for worker and ingress health.

Platform controls#

Operations › Platform controls manages:

• Platform availability: maintenance mode with a banner and schedule. Super admins keep access; other operators can be allowlisted.
• Organisation maintenance for a single organisation.
• Global defaults: quiet hours, daily and weekly caps, request rate limit, second-approver threshold, registration mode, session lifetime, minimum password length, supported networks, platform name and tagline, default SMS price, support contact and legal notice.
• Sign-in policy: which sign-in methods are enabled.

Kill switches#

Operations › Kill switches stops new SMS submissions for the platform, a provider or a tenant. Use it in an incident, then investigate. Resume from the same page.

Audit#

Operations › Audit searches the operator and tenant audit chains. Entries are hash-linked; retention is at least five years, and purges leave a signed marker so the chain still verifies. Changing the retention period needs two operators.

Cross-tenant view and records#

Operations › Cross-tenant view gives an inventory across organisations for platform oversight, without contact values. Manage records handles operator-level corrections with reasons.

Tenant notices#

Operations › Tenant notices publishes a banner and account notification to an organisation, with a priority (information, warning or critical), start and end times and an optional email copy to the owner. Archive or restore notices with a reason.

Support access#

Operator support reaches a tenant only through a time-boxed, reason-logged grant approved by a super admin. The support view is read-only and ends after 15 minutes or when the grant expires.

Incident response#

If you suspect a breach:

1. Contain it (kill switches, session revocation, credential rotation).
2. Record it in the breach register.
3. Notify the ODPC within 72 hours of becoming aware, as the Data Protection Act requires, and affected people where required.
4. Notify the tenants whose data is involved.