We already have an IT provider — does this work alongside?

Usually yes. It is often about something they do not cover: a virtualisation cluster, mail security, logging, or a second opinion before a larger investment. We do not automatically take over everything.

What does highly available actually mean?

Two or more nodes with synchronously mirrored storage: if one server fails, the virtual machines carry on somewhere else. Ours runs across two sites — the same architecture we would set up for you.

Do you also run it day to day?

Yes: monitoring, updates, backups and restore tests. Without a 24/7 promise we could not keep — what we commit to is a response time we actually hold.

Whichever adds up. Your own hardware pays off with steady baseline load and large volumes of data; the cloud pays off with fluctuating demand and little capital tied up. Often the mix is right: production in-house, backup and test environments outside.

05 · IT

General IT Solutions

Nobody notices infrastructure while it works. It gets noticed on the day it fails — and that is when you find out whether backup, network segmentation and monitoring were more than a tick on a list.

The foundation everything runs on: high-availability infrastructure, security and operations — planned, set up and maintained for the long run.

›Virtualization & HA with Proxmox, LINBIT DRBD/LINSTOR
›Hosting on Hetzner & own hardware, MinIO object storage
›Security: firewalling, CrowdSec, monitoring with Graylog
›Mail servers, backups, maintenance and updates

ProxmoxHetznerLINBITMinIOGraylogCrowdSec

Who this is for

Businesses that have outgrown the single-server model, and those who want an existing environment taken over without rebuilding everything. Also as a second opinion on a quote you already have.

What we run ourselves

The infrastructure of a corporate group with nine companies and three locations—for twenty years, without a contract with an external data center.

Virtualization across two fire compartments: A cluster of six nodes, distributed across two fire compartments, with mirrored storage instead of a central storage device. If one compartment fails, the machines in the other continue to run.
In-house mail gateway: Filtering, quarantine, and mailboxes are self-managed, with proprietary scanning modules in place. No mailboxes hosted by a provider that monitors the content.
Attack detection: Centralized log analysis with custom rules and alerting. Anomalies are reported as alerts, not stored in a file that no one reads.
Data Backup with Verification: Daily backup with automatic verification to ensure it can actually be restored. An unverified backup is no backup at all.
Triple DNS Resolution: Three dedicated DNS servers with filtering. If one fails, no one notices.

How a project runs

We don’t rebuild anything until we know what’s there. The most common finding in the inventory is not missing technology, but a data backup that has never been restored.

1Inventory
What’s running where, what depends on what, where are the licenses, when does support expire. The result is a list of risks, sorted by impact.
1–2 days
2First things first
Check data backups, close open access points, and apply updates. These steps cost little and eliminate most of the risk.
1–2 weeks
3System overhaul
Virtualization, storage, network, email, monitoring—in steps that can be rolled back individually. Changes are made during maintenance windows, not on Tuesday mornings.
Depending on scope
4Operation or handover
We provide ongoing support, or we hand things over to your team—with sufficient documentation. Both options are fine, just avoid the middle ground where no one takes responsibility.
Ongoing or one-time

What it costs

Operating your own infrastructure is rarely cheaper than renting, if you do the math honestly—but it’s predictable, and it keeps the data in-house. We calculate both options, including labor costs.

What’s Often Overlooked: Electricity, maintenance windows, and labor hours. A comparison that pits only the hardware against the monthly rent always favors the hardware—and yet it’s still inaccurate.
What Almost Always Pays Off: The cost of user licenses, which are billed per person per month. Once you reach a certain number of employees, this is the largest single expense item.
Inventory assessment: At a fixed price, regardless of whether we continue to work together afterward. You’ll receive the list in any case—even if you give it to someone else.

What you get

✓A risk list sorted by impact, not by effort
✓High availability without a central storage device as a single point of failure
✓Data backups whose restorability is verified rather than simply assumed
✓Documentation that survives a change in service provider
✓Preparatory work for NIS2—evidence that will be required anyway

Frequently asked

We already have an IT provider — does this work alongside?: Usually yes. It is often about something they do not cover: a virtualisation cluster, mail security, logging, or a second opinion before a larger investment. We do not automatically take over everything.
What does highly available actually mean?: Two or more nodes with synchronously mirrored storage: if one server fails, the virtual machines carry on somewhere else. Ours runs across two sites — the same architecture we would set up for you.
Do you also run it day to day?: Yes: monitoring, updates, backups and restore tests. Without a 24/7 promise we could not keep — what we commit to is a response time we actually hold.
On-prem or cloud?: Whichever adds up. Your own hardware pays off with steady baseline load and large volumes of data; the cloud pays off with fluctuating demand and little capital tied up. Often the mix is right: production in-house, backup and test environments outside.

Let's talk. Services