Service Level Terms
Our availability targets, what happens when we miss them, and an honest account of where a young platform stands.
1. Where we actually stand
We are not selling you five nines
Naijacloud has been running production workloads since 2025. We do not yet have the years of operating history, the regional redundancy or the dual-region database failover that a four- or five-nines commitment would require. Publishing one now would be a marketing number, not an engineering one.
So this document gives you three things instead: the targets we hold ourselves to today, the credits you get when we miss them, and everything we publish so you can check our work rather than take it on faith. As redundancy lands, the targets here go up, and we will say so in the changelog.
These terms apply to paid Pro and Scale workloads. The free and Hobby tiers are best-effort with no credits, and beta features are excluded on every plan.
2. Monthly availability targets
Component
Target
Allowed downtime
Notes
Web services & static sites (Pro, Scale)
99.9%
≈43m / month
Per service, per region.
Object storage
99.9%
≈43m / month
Availability of the API, in-region.
Managed databases (single instance)
99.5%
≈3h 40m / month
Single-instance Postgres. Recovery involves a restart or restore.
Managed databases (high availability)
—
—
Standby replicas ship H1 2027. Until then we will not quote a higher number.
Cron jobs
Best effort
—
Jobs are retried; start times may drift under load.
Dashboard, CLI & API
99.5%
≈3h 40m / month
Control plane. Running services keep serving traffic if it is down.
Availability is measured per calendar month from our external monitoring, as the percentage of one-minute intervals in which a component served requests successfully. The measurement we publish on the status page is the one we use for credits.
3. Service credits
If we miss a target for a component you pay for, you can claim a credit against the affected component’s fees for that month:
Monthly availability
Credit (99.9% target)
Credit (99.5% target)
Below target, above 99.0%
10%
10%
99.0% – 95.0%
25%
25%
Below 95.0%
50%
50%
Below 90.0%
100%
100%
Credits are applied to your next invoice, are capped at 100% of the affected component’s monthly fee, and are the sole and exclusive remedy for missed availability. They are not refunds and are not payable in cash.
4. What does not count
- Scheduled maintenance announced at least 72 hours ahead on the status page. We aim for under 4 hours a quarter, in the 01:00–05:00 WAT window.
- Emergency maintenance to patch an active security vulnerability, announced as soon as we can.
- Anything caused by your own code, configuration, migrations, exhausted plan limits or unpaid invoices.
- Beta, preview and experimental features, and the free and Hobby tiers.
- Failures of your Git provider, DNS registrar, payment provider or other third parties outside our control.
- Force majeure, including national network or power failures beyond our facilities, and government action.
5. Claiming a credit
Email [email protected] within 30 days of the end of the affected month with your workspace name, the affected service, and the approximate dates and times. We will confirm against our monitoring and reply within 10 business days. You do not need to prove the outage. We will look it up.
6. What we publish either way
- A live status page with per-region health for every component, and 90-day uptime history.
- An incident entry opened within 15 minutes of us confirming customer impact, updated at least every 30 minutes until resolved.
- A written postmortem within 5 business days of any incident over 30 minutes: timeline, root cause, and what we are changing. No blame, no vagueness.
- Advance notice of maintenance windows, and of any change to the targets on this page.
Questions about this policy?
We answer legal and privacy mail within five business days.
Naijacloud Technologies Ltd · 14 Aba Road, Port Harcourt, Rivers State, Nigeria. Previous versions of every policy are archived and available on request.