Service level agreement

This applies to managed hosting, managed mail and IT support plans. It forms part of the agreement for those services, and where a signed agreement says something different, the signed agreement wins. Project work is governed by its own quotation and scope rather than by this document.

The uptime commitment

Managed hosting is committed to 99.9% availability, measured monthly. Over a calendar month that allows roughly 43 minutes of unavailability. Availability is measured from outside our infrastructure, by external monitoring at one-minute intervals, and a service is unavailable when two consecutive checks from separate locations fail.

Planned maintenance

Maintenance that requires downtime is scheduled outside 07:00 to 17:00 on business days, announced at least 48 hours in advance in writing, and kept inside the announced window. Planned maintenance inside those rules does not count against the 99.9%. Emergency patching for an actively exploited vulnerability may happen with less notice, and we tell you what we did as soon as it is done.

How quickly we respond

Support hours are Monday to Friday, 07:00–17:00 SAST. Managed plans reach a person outside those hours for anything at severity one. Response means a human being has read it, understood it and told you what is happening, not that an automatic acknowledgement has gone out.

• Severity one, a service is down or a business cannot work: response within one hour in support hours, and out of hours on a managed plan. Worked continuously until it is resolved or downgraded.
• Severity two, something important is broken but there is a way around it: response within four business hours.
• Severity three, a fault that is not blocking anybody: response by the end of the next business day.
• Severity four, a question, a change request or a piece of advice: response within two business days.

Backups and restoring

Managed hosting is backed up nightly and backups are kept for thirty days. Restores are tested on a schedule rather than assumed to work, because a backup that has never been restored is a hope and not a backup. A restore from the most recent backup is started within four business hours of your asking for it, at no charge.

What is excluded

• Anything caused by your own change to a system you administer, or by software you installed on it.
• A failure at a third party we do not control, including your connectivity, your domain registrar, and a provider outage that takes out a whole region.
• Denial of service attacks and other deliberate attacks by third parties, beyond the mitigation the platform already provides.
• Suspension under the acceptable use policy, or suspension for non-payment after written notice.
• Force majeure, including sustained loss of grid power beyond the runtime of the provider's own backup arrangements.

What happens when we miss

If availability falls below 99.9% in a calendar month, you get a credit of 10% of that month's fee for that service, and 25% if it falls below 99%. Claim it by writing to us within thirty days of the month end and we will check it against the monitoring rather than argue about it. Credits go against the next invoice. If we miss the commitment in three consecutive months you may cancel the service immediately, without penalty, and we will help you move.

When something breaks

One of us takes the incident and stays on it. You get plain-language updates saying what broke, what it affects and when the next update is coming. Anything lasting more than thirty minutes gets a short written account afterwards: cause, fix, and what changes so that it does not repeat. The status page sets all of this out in more detail.

Leaving

Recurring services run month to month unless the agreement says otherwise, and either of us may end them on one calendar month's written notice. On the way out you get a full export: files, databases, DNS records, mailbox data and the credentials to everything in your own name. We do not charge for it and we do not slow it down.