A managed server you do not have to watch yourself

Your website or application runs somewhere, and as long as everything works you do not think about it. Until a disk fills up, a certificate expires on a Friday afternoon, or an update knocks everything over. Then it turns out nobody knows exactly whose server that actually is. The hosting provider delivers the machine, your developer delivers the code, and the layer in between belongs to nobody.

What management actually includes

Managed hosting means we take over that layer in between. The machine itself sits in our managed environments within the EU. What we do is everything that runs on top: the operating system, the web server, the database, the certificates, the firewall, and the backups.

That does not become an open end. The management agreement states which parts we keep up, which ones you or your own vendor keep, and what happens if something falls outside that. So in an outage you know right away whose move it is, instead of three parties pointing at each other while your website is down. If you are still choosing between a shared machine and a box of your own, we simply take a look with you in the intake: sometimes VPS hosting is enough, sometimes Dedicated servers are the logical step.

Updates and maintenance without surprises

The biggest risk on a server is not an attack from outside, but overdue maintenance. Packages that have not been updated in two years, a PHP version that no longer gets security updates, a certificate nobody renews. We keep that up at a fixed rhythm: small security updates go along, big version jumps we plan in consultation and first test on a copy of your environment.

If something does break in such a jump, we roll back to the previous state instead of tinkering live. You hear beforehand when we do something and what the noticeable effect can be, even if that effect is zero. Maintenance you only notice when it goes wrong is maintenance that was planned too late.

Monitoring and the agreement on who watches

With us, monitoring is a system. Measuring points continuously watch disk space, memory, processor load, the run time of your background tasks, and the validity of your certificates. If something deviates from the value we have together called normal, an alert goes out along an escalation line agreed in advance: first a signal, then a person, and you know which person that is.

A single server needs recovery time when it fails, however well it is kept up. If you cannot afford that time, that is a design question and not a management question; High availability and failover then spreads your environment over several machines. We tell you that right away in the intake, even if it means the simple solution is not enough.

Backups, recovery, and ownership

A backup that has never been restored is an assumption, not a backup. We set the copies up so that they sit in a different place than the server itself, with a retention period that fits what you keep, and we really restore one once to see how long it takes. That recovery time is the number you can use. For heavier scenarios, where a whole data center drops out or ransomware runs through your files, Backup and disaster recovery belongs with it.

What you place with us stays yours. Your data sits in the EU, the code and the data remain your property, and you get the documentation of the setup and the access to your own environment. If you switch to another party or go do it yourself, you take that setup with you; we build in nothing that only works with us. Every collaboration starts with a free intake in which we first look at what you have running now, before we propose anything.