What we commit to, in plain words
The availability commitment on our Cloud VPS, the credit it pays when we miss it, what is excluded, what the second disk copy does and does not do, what a plain VPS carries, and the fair-use rule on unmetered bandwidth.
Cloud VPS: 99.9% monthly availability
Every Cloud VPS carries a 99.9% monthly availability commitment. When we fall short of it, you are credited — not as a goodwill gesture, but at the rate set out here.
- What is measured
- Whether the host running your server is reachable from our network, measured per calendar month. Downtime is counted from the moment we detect it or you report it, whichever comes first, until the host is reachable again. It is the host that is measured — what runs inside your server is yours, and is covered under the exclusions below.
- The credit
- 2× the pro-rata price of the shortfall — the share of that server’s monthly fee that the downtime beyond the allowance represents — capped at one month’s fee for that server. It is applied to your account as credit against your next invoice.
- How to claim
- Open a support ticket within 30 days of the end of the month in question, naming the server and the period you believe it was unreachable. We check it against our own records and apply the credit.
Worked example
A Cloud Professional at £50.99 a month, in a 30-day month, unreachable for 7 hours.
- Hours in the month
- 720 hours
- Downtime the commitment allows at 99.9%
- 43 minutes
- Downtime over the allowance
- 6.3 hours
- Share of the month
- 0.9%
- Pro-rata price of the shortfall (0.9% of £50.99)
- £0.45
- Credit at 2×
- £0.90
The cap is one month’s fee for that server — £50.99 here — so the credit stands at £0.90. A longer outage earns a larger credit up to that cap; a month with no more than 43 minutes of downtime earns none.
The price shown is the price you pay — no VAT is added.
What is excluded
Time in any of the following is not counted as downtime against the commitment:
- Scheduled maintenance announced at least 48 hours ahead. Planned maintenance normally moves your server live to the other host without a reboot; the notice is for the cases where a reboot is needed.
- Faults inside your own operating system, software or configuration. The host is ours; what runs on your server is yours. A crashed application, a full disk, a misconfigured firewall or a kernel you upgraded is not host downtime, however unreachable the server may be.
- Events outside our control — the upstream network, the power supply to the data centre, and force majeure.
- Suspension for non-payment or for abuse of the service, including a breach of the fair-use rule below after notice.
What the second disk copy does and does not do
A Cloud VPS is the same server as our VPS, except that a second copy of its disk is kept on another physical host in our Derby data centre, refreshed every 5 minutes. That is the whole mechanism, and it is worth being exact about what it covers.
What it does
- If the host running your server fails, your server restarts on the other host automatically — same address, same files as of the last copy — typically within minutes, instead of waiting for an engineer to repair the hardware.
- Because the copy is refreshed every 5 minutes, up to 5 minutes of writes can be lost in a host failure. Anything written after the last copy is not on the other host.
- Planned maintenance uses the same second copy to move your server live to the other host, normally without a reboot.
What it does not do
- It is not a backup. The copy mirrors the disk as it is — so a deleted file, a ransomware encryption or a bad deploy is copied across just as faithfully as everything else, within 5 minutes.
- For those, every Cloud VPS includes 7 daily restore points, kept on the other host, from which you can roll the server back by ticket. They are the thing that covers deletion, ransomware and application faults; the second copy is not.
- Restore points are kept in the same data centre as the server. They are not a substitute for your own off-site backups of anything you cannot afford to lose.
Plain VPS
A plain VPS carries no availability percentage and no credit. It runs on one host, with no second copy of its disk elsewhere.
If the hardware under it fails, we repair or replace the failed hardware and bring your server back on it. If you have the daily snapshot add-on (£2.20 a month), we restore from your most recent snapshot; otherwise the server is restored from your own backups. How long that takes depends on the fault, and we do not print a figure we cannot keep.
If a short outage is not acceptable to you, the Cloud VPS is the product with the commitment above.
Fair use
Bandwidth on our VPS and Cloud VPS is unmetered on a shared 1 Gbit/s port. There is no monthly transfer allowance and no overage charge.
Fair use means that sustained saturation of the port, or traffic patterns that degrade the service for other customers on it — floods, bulk mail, scanning — may be rate-limited after we have given you notice.
Nothing here caps normal web, mail, database or backup traffic. A busy site, a nightly backup to somewhere else, a large download now and then — that is what the port is for.
Support
Support is UK-based, in UK business hours, by ticket. Credit claims, restore-point rollbacks and anything else on this page start with a support ticket.
The company behind the service, and its registered details, are stated in the footer of every page on this site.