Uptime monitoring
Continuous availability checks per site, surfaced against the site's own history so a blip and a real outage look different. Uptime state is a first-class signal the platform acts on, not a graph nobody reads.
Uptime, performance & error monitoring
Know your site is up before anyone tells you
Zinn Digital® watches every site you host with us — uptime, Core Web Vitals, errors, traffic and resource usage against your entitlement — and surfaces it per site in the dashboard. Alerts go to email, Telegram, Slack or your own webhook, and our status page lives off-platform so it still answers when we cannot.
- 99.99%uptime assurance
- 4alert channels: email, Telegram, Slack, webhook
- Off-platformindependent status page
- 100,000+PBN sites hosted across every niche
Four signals per site, in one place
Hosting dashboards usually tell you what you are paying for. Ours tells you what your site is doing. Every site you run with us reports the same set of signals, per site, with no agent to install and nothing to wire up.
Core Web Vitals & performance
Real performance data per site — the metrics Google actually ranks on. Slow pages show up as a measurement you can act on, next to the caching controls that fix them.
Error tracking
HTTP and PHP errors captured per site, so a fatal from a bad plugin update or a broken deploy surfaces as an error rate rather than as a customer complaint days later.
Traffic, bandwidth & resource usage
Bandwidth and traffic per site, plus resource usage measured against your plan entitlement — including per-site LVE statistics — so you can see which site is working hardest and whether it is near a ceiling.
Alerts that reach you where you already work
A dashboard you have to remember to open is not monitoring. Every signal feeds the platform's notification layer, and you choose where it lands.
- Four channel adapters, live: email, Telegram, Slack and your own webhook endpoint — each behind a swappable adapter, so no single vendor owns your alerting.
- Preferences are set per organisation and per user: decide which events go to which channels, so the person on call gets the outage and the finance contact does not.
- Webhooks are HMAC-signed, retried with backoff and logged — every delivery is inspectable and replayable from the dashboard, so a receiver that was down does not mean an alert you never saw.
- Every signal rides the same durable event bus as the rest of the platform. A consumer that lags never blocks the producer, and no event is silently dropped.
- Alerts are localisable like everything else we send — 58 languages, in the language the recipient set.
A status page that survives the outage
A status page hosted on the infrastructure it monitors is useless at exactly the moment you need it. Ours is deliberately somewhere else.
Our public status page is self-hosted on separate infrastructure, entirely off-platform — a different box, a different failure domain from the control plane and the fleet it reports on. When something goes wrong with the platform, the status page keeps answering, because nothing it depends on is inside the incident.
It is branded, fed by the same monitoring that drives your per-site alerts, and independent by design rather than by accident. The public frontend follows the same principle: the marketing site and help centre are deployed separately from the control-plane cluster, so a platform incident does not take our documentation down with it.
This is the same reasoning that runs through the rest of the platform — separate failure domains, no shared fate, nothing that can take out both the thing and the tool you use to diagnose it.
Monitoring your AI assistant can actually read
Signals are worth more when something can act on them. Ours are wired into the assistant and into the MCP server, not locked in a graph.
Pre-ticket diagnostics
When you start to raise a ticket, the assistant runs a live diagnostic on the affected site first — uptime, HTTP and PHP errors, SSL and DNS state, resource limits and LVE faults, malware flags, plugin conflicts, database health and cache — before a human is ever involved.
Safe auto-fixes, then re-check
Within your permissions and the platform's guardrails it will clear cache, restart PHP workers, repair or renew SSL, repair database tables, disable a fatally-erroring plugin or purge the CDN — then re-check the signal and confirm whether it actually resolved.
Escalation with the evidence attached
If a human is still needed, the ticket arrives pre-filled with the diagnostics, what was tried and what happened — so support starts from the facts instead of asking you to describe the symptom.
Readable over MCP
Connect any MCP-capable AI tool and it can read site health, recent logs, metrics, uptime and errors as resources — so your own agent can diagnose before it acts, under OAuth 2.1 with org and role-scoped, revocable tokens and a full audit log.
Core Web Vitals as an SEO signal, not a vanity chart
Performance monitoring belongs next to the tools that change performance. On our platform it sits between the caching stack and the SEO layer.
- Every plan carries the free SEO Basics tile: is the site indexed, its Authority Score, and homepage Core Web Vitals — the three numbers that tell you whether a site is working as an SEO asset.
- Footprint-Free plans add automatic Google indexing checks with weekly monitoring, so an indexation problem surfaces as an alert rather than as a traffic drop you notice a month later.
- When CWV degrades, the fix is one page away: LiteSpeed Enterprise with LSCache full-page caching, per-site Redis object cache and CDN edge delivery, coordinated with smart auto-purge.
- Deeper SEO monitoring is available as paid add-ons — rank tracking, backlink monitoring, competitor intel and a monthly SEO health audit — priced separately from your hosting plan.
- Hands-on site speed optimisation is available as a purchasable extra when you would rather we did the tuning than you did.
Built to watch a million sites
Behind your dashboard is the same observability stack we run the platform on: Prometheus and Grafana for metrics, OpenTelemetry for distributed tracing across the engine, workflows and drivers, and centralised structured logs that are tenant-tagged with audit data kept separate.
Metrics are collected per service, per worker and per tenant, with per-site LVE statistics from the isolation layer. Every workflow run — provisioning, migration, enforcement — is visible end to end, which is why a support answer can be specific instead of a guess.
Monitoring at scale is a data-volume problem, so it is engineered as one. High-volume metrics are time-partitioned and rolled up rather than kept per-site per-minute forever, traces are sampled, and per-tenant cardinality controls stop one noisy account degrading everyone else's dashboards. The design target is a million sites, not the number we run today.
Alerting is routed through Alertmanager to our on-call rota with SLOs defined and tracked — kept deliberately separate from the customer notifications above, so operational noise never lands in your inbox and your alerts never queue behind ours.
FAQ
What exactly do you monitor on my sites?
Per site: uptime, Core Web Vitals and performance, HTTP and PHP error tracking, traffic and bandwidth, and resource usage measured against your plan entitlement — including per-site LVE statistics from the isolation layer. It is all surfaced in the dashboard per site, with no agent to install.
How do I get told when a site goes down?
Through whichever channels you connect: email, Telegram, Slack or your own webhook endpoint. Notification preferences are set per organisation and per user, so you decide which events reach which channel and which person. Webhook deliveries are HMAC-signed, retried with backoff and logged, so you can inspect or replay any delivery.
Does the status page go down when the platform does?
That is exactly what it is designed to avoid. The status page is self-hosted on separate infrastructure, entirely off-platform, in a different failure domain from the control plane and the fleet it reports on — because a status page running on the infrastructure it monitors is useless during an outage.
What does the 99.99% uptime assurance mean?
It is the uptime target we build and operate to, and it is stated as an assurance on Footprint-Free plans. It is a commitment about how the platform is engineered — isolated per-site compute, horizontally scalable stateless services, separate failure domains — rather than a published measurement of past availability. If you need a contractual service level, ask us and we will scope it.
Can my own tools or AI agents read the monitoring data?
Yes. Site health, recent logs, metrics, uptime and error data are exposed as resources over our hosted MCP server, so any MCP-capable AI tool — Claude, Cursor, ChatGPT and others — can read them and diagnose before acting. Access is OAuth 2.1, scoped to your organisation and the connecting user's role, revocable per tool, and every action is audit-logged. Customer webhooks cover the same ground for conventional integrations.
Is monitoring extra, or part of hosting?
The observability signals described here are part of the platform and surfaced in your dashboard for the sites you host with us, and the free SEO Basics tile — indexed, Authority Score, homepage Core Web Vitals — is bundled on every plan. Deeper SEO monitoring such as rank tracking and backlink monitoring is sold as separate paid add-ons, and hands-on speed optimisation is a purchasable extra.
Can I try it before I commit?
Yes. The Footprint-Free trial is 7 days, card-free and covers up to 5 sites — no payment details, no commitment. Deploy a few sites and watch what the dashboard actually reports before you decide.
Related
Stop finding out from your customers
Start a card-free 7-day trial and see uptime, Core Web Vitals, errors and resource usage per site from the first deploy — with alerts wherever you want them. Free migrations and a 30-day money-back guarantee behind everything.
Start free