Infrastructure work is invisible when it is right and costly when it is not. The failure is rarely a dramatic outage. More often it is a deploy nobody dares run, a server only one person understands, or a backup that has never been restored.
What is delivered
- A hosting design that fits the system: what runs where, how traffic reaches it and what happens when one part fails.
- Reproducible deployments. Containerised and running the same way on every environment, so a deploy is a non-event.
- Security as configuration. TLS, secrets kept out of the repository, least-privilege access and backups that have been restored at least once.
- Monitoring worth having. Alerts on conditions that need a human, not a wall of dashboards nobody opens.
- Documentation and the configuration as code, so the setup can be rebuilt from the repository.
How it works
We start with a short review of how the system runs today. Then we write down a fixed scope and price, build the setup in milestones you can check, and hand over the configuration and the documentation. If you want us to keep operating it, we do. If you would rather run it yourselves, the documentation is written for that.
Work to look at
- Our own products. We run five products in production (an ERP, an EMS and SCADA platform, a ticketing system, a CRM and an invoicing platform) behind a shared reverse proxy on infrastructure we manage. See the energy platform and the ERP.
- Monitoring on Azure Stack Hub. An enterprise monitoring platform built for a Microsoft cloud partner, with Python exporters, Prometheus and Grafana.
- An OpenStack private cloud lab, built to learn how the platform is installed and run.
The last two are listed in our founder’s portfolio.
Related services
Pipelines, container orchestration and release automation are covered on the DevOps consulting page. If the work is mainly about connecting systems, see software integration.