Three kinds of bad day
Not every problem is a disaster. Most are small, and they are handled long before they reach you.
A change goes wrong
A new version starts beside the old one and only takes over once it is healthy. If it is not, the old one keeps serving and the new one is rolled back by itself. Most visitors never notice.
Something gets deleted
Deleted the wrong thing on Friday afternoon? Happens. Your system and data are backed up every day, kept separately from the live site, so it is still there.
The whole server is lost
A hardware fault or a data centre outage. The rarest of the three, and the one this page is mostly about.
If the server is lost completely
The recovery is prepared in advance and rebuilds the entire server from the daily backups.
The moment it happens
We know
Monitoring checks the sites continuously and alerts us immediately when they stop answering.
Recovery starts
A prepared procedure
A recovery procedure prepared and rehearsed in advance starts. Nobody has to improvise.
Restore
A new server
A new server is created, and every client’s systems, databases and files are restored from the latest daily backup.
Within 30 minutes
Back online
The whole system is running again and visitors can reach the sites once more.
What makes it possible
Thirty minutes is not luck. It rests on work that happens every day, whether or not anything goes wrong.
Daily backups, elsewhere
Every server and all stored files are backed up daily with a separate provider, so a problem at one cannot take the copies with it.
A daily snapshot
The state of the whole machine is captured every day as well, not only the data on it.
Protected files stay put
Private files are kept in separate storage, not on the server itself. A lost server does not touch them, and the new one connects to the same storage.
Monitoring that wakes us up
Downtime raises an alert straight away, so recovery starts with us, not with a client’s phone call.
Where it stops
Worth knowing before you rely on it.
Backups are daily
After a total failure, data created since the last daily backup is not yet in that backup. If your system cannot afford to lose a day, tell us, and we plan for it.
A shared server goes down together
Most clients share one managed server, each in their own isolated space. An outage there affects all of them until recovery. A server of your own is possible if you need it.
In development
A static fallback while the server is down
We are making static copies of the pages that don’t need a database. During a recovery they would stay browsable. Forms, logins and anything else that needs the database would wait until the server is back.
What would a lost day cost your business?
Tell us how your system is used, and we’ll tell you honestly whether daily backups are enough for it.