System administration

A common situation. We start with an assessment: we find every server and service, work out what does what and produce a diagram. That takes one to four weeks depending on size. After that we administer from the documentation.

Before updating we take a snapshot or a backup, roll out to a test environment first if there is one, and do the work in an agreed window, usually at night or at the weekend. If something goes wrong, we roll back before the working day starts.

With both, including the Russian Linux distributions used for import substitution. A mixed setup, with the domain on Windows and some services on Linux, is normal for us.

If the company has many internal systems and constant change, an in-house specialist is valuable: they know the business better. In that case we take the infrastructure, the on-duty cover and the narrow specialisms, while they handle application-level work.