Service Level Agreement
Service Level Agreement
Last updated: 16 September 2026
This Service Level Agreement states what availability LBREEZE LIMITED commits to, how we measure it, what is excluded, and what you get if we miss it. It applies to services whose plan says an SLA is included, and it forms part of our Terms of Service.
An SLA is a promise with a number attached. Everything below is written so that either side can work out, from records, whether the promise was kept.
1. What is covered
This SLA covers the availability of the infrastructure we operate: the virtual machines and Incus instances on your plan, the network to the edge of our infrastructure, the storage those instances depend on, and the lbreeze customer portal.
It does not cover the availability of software running inside your instances, whether you installed it or our automation deployed it. A virtual machine that is powered on, reachable and serving traffic at the network layer is available under this SLA even if an application on it is misconfigured or has crashed. Responsibility for the layers above is set out in the Shared Responsibility Statement.
2. The availability commitment
We commit to 99.9% availability per calendar month for each covered service, measured as set out in section 3. [TO CONFIRM: 99.9% allows roughly 43 minutes of downtime per 30-day month. Confirm this is the figure you want to be held to, and mirror it in Settings so the platform and this document cannot drift apart.]
Availability is calculated per service, not across your account. One instance being down does not put another instance in breach, and it does not matter how many services you run.
3. How availability is measured
For each covered service and each calendar month:
Availability % = ((Total minutes in the month − Excluded minutes − Unavailable minutes) ÷ (Total minutes in the month − Excluded minutes)) × 100
A minute is unavailable if, for that whole minute, the service does not respond to our monitoring from at least two independent monitoring locations. Partial-minute interruptions are not counted. Degraded performance is not unavailability unless it makes the service unusable, in which case we will treat it as unavailability and say so.
Our monitoring records are the primary evidence. If you have your own monitoring that disagrees, send it with your claim: where the two conflict and we cannot reconcile them, we resolve the difference in your favour.
4. What is excluded
Excluded minutes do not count against the commitment, and they are removed from the denominator so they neither help nor harm the percentage:
- Scheduled maintenance, announced at least [TO CONFIRM: notice period in hours, commonly 72] in advance by email and on the status page. We schedule it inside a standard window of [TO CONFIRM: your maintenance window, for example Sunday 02:00 to 06:00 UK time] and keep it as short as we can.
- Emergency maintenance needed to close a security vulnerability or prevent imminent failure. We give as much notice as the situation allows, and we publish what happened afterwards.
Unavailability caused by any of the following is not our failure and does not count:
- your own configuration, software, updates or resource exhaustion inside the instance;
- suspension under the Terms of Service or the Acceptable Use Policy, including suspension for non-payment;
- an attack on or through your service, including a distributed denial of service attack directed at your workload, except where our own mitigation was available and we failed to apply it;
- your failure to follow a documented instruction or to act on a notice we gave you;
- anything beyond our reasonable control as described in the Terms of Service, including upstream network, power or data centre failures at a level we have no contractual remedy against;
- beta, preview, free or trial services, which carry no SLA.
5. Service credits
Where we miss the commitment in a calendar month, you can claim a credit against the monthly fee for the affected service:
| Availability achieved in the month | Credit |
|---|---|
| 99.0% up to but not including 99.9% | 10% of that month's fee |
| 95.0% up to but not including 99.0% | 25% of that month's fee |
| 90.0% up to but not including 95.0% | 50% of that month's fee |
| Below 90.0% | 100% of that month's fee |
[TO CONFIRM: this ladder is a common market shape, not a decision you have made. Confirm it, and confirm the cap in the next paragraph, before publishing.]
Credits in any month are capped at 100% of that month's fee for the affected service, are applied against future invoices, and are not paid in cash. Where a service is on an annual plan, the monthly fee is one twelfth of the annual fee. Credits are the sole and exclusive remedy for missed availability, except that if we miss the commitment in three consecutive months you may terminate the affected service immediately and receive a pro-rata refund of prepaid fees.
6. Claiming a credit
Open a ticket with billing@lbreeze.com within 30 days of the end of the month in question, with the service identifier, the dates and times of the unavailability including the time zone, and any monitoring evidence you have. We will respond within 10 working days with the calculation and the credit, or with the reason the claim does not qualify.
We do not apply credits automatically. That is deliberate: an automatic credit calculated only from our own monitoring would silently miss anything our monitoring did not see, and a claim gives you the chance to put your own evidence in front of us. If you would rather we monitored a specific endpoint of yours as the measurement point, ask and we will set that up.
7. Support
| Channel | Availability |
|---|---|
| Ticket and email | Monitored [TO CONFIRM: your standard support hours, for example Monday to Friday, 09:00 to 17:00 UK time, excluding English public holidays] |
| Emergency escalation for critical incidents | 24 hours a day, every day, via [TO CONFIRM: the escalation route customers should use out of hours] |
Target first response times, measured from when a ticket is opened and within the hours above, except for critical incidents which are measured around the clock:
| Severity | Definition | First response target |
|---|---|---|
| Critical | Service wholly unavailable, or a confirmed security incident affecting it | 1 hour |
| High | Major function unusable, or serious degradation, with no workaround | 4 hours |
| Medium | A function is impaired, or there is a workaround | 1 working day |
| Low | Question, request, or cosmetic issue | 3 working days |
These are response targets, not resolution targets: a first response means a person has picked the ticket up and told you what is happening, not that the problem is fixed. We do not commit to a resolution time, because the honest answer depends on the cause, but we will give you an assessment and keep you updated at intervals we agree on the ticket. Missing a response target does not by itself generate a service credit.
Severity is set by us on the facts. If you disagree, say so on the ticket and we will re-assess immediately.
8. Status and incident communication
Service status is published at [TO CONFIRM: your status page URL]. For any incident affecting more than one customer, we post an initial notice as soon as we have confirmed it, update it at least hourly while it is open, and post a closing note when it is resolved. For a critical incident lasting more than four hours we publish a post-incident review within 10 working days, covering what happened, why, and what we are changing.
9. Backups
Backups are an optional service. Where your plan does not include them, there is no backup of your data and this SLA gives you no restoration right. Where your plan does include them, the frequency, retention and restore target are stated on the plan, and a failure to hold a backup we were contractually required to hold is a breach of this SLA rather than a missed-availability event; the remedy is in the Terms of Service.
Keep your own copy of anything you cannot afford to lose. This is not a disclaimer dressed as advice: restoring from a backup is a recovery process with a duration, and for some data that duration is too long whatever we commit to.
10. Changes to this SLA
We may change this SLA. Where a change reduces a commitment, we will give at least 30 days' notice by email, and you may terminate the affected service before it takes effect and receive a pro-rata refund of prepaid fees. Improvements take effect when published. The current version is always at lbreeze.com/legal/sla with the date it was last updated.