Area 03 · Infrastructure and support

System administration

Equipment does not need a rescuer arriving in the middle of the night. It needs patches applied on time, drives that are not full and someone reading the event logs before anything fails. That unglamorous work is what we take on.

8
services in this area
24/7
server monitoring
Windows
and Linux in one team
1 month
to take over the infrastructure

What we do on a schedule

Failures are prevented by the dullest part of maintenance. That is why it runs strictly to a schedule rather than when someone gets round to it.

Logs and service status

Every day. Stops you missing a failed component you would otherwise learn about a week later.

Free space and load

Every day. Stops the database halting because a drive has filled up.

Restoring the archive as a test

Once a month. Removes the main danger: a backup you cannot restore from.

Security patches

Once a month. Closes the holes that have become common knowledge, attackers included.

Reviewing accounts and permissions

Once a quarter. Removes access from people who left the organisation long ago.

First we document the estate and only then take it on. The diagram, the list of services, where the data sits and the chains of dependencies come together in a document that is yours for good. Without it all the knowledge lives in one specialist's head and leaves with them.

Questions and answers

That is how it is almost everywhere. We start with discovery: finding every node and service, establishing who is responsible for what, drawing the diagram. It takes from a week to a month depending on scale. After that the work follows the document rather than guesswork.

Before installing anything we take a snapshot; if there is a separate test rig we try it there first, and the work itself is done at an agreed time, usually at night or at the weekend. If it goes wrong, the rollback to the previous configuration is finished before the working day starts.

Both platforms, including the local Linux builds used when migrating to trusted software. A mixed setup where the directory service runs on Windows and the application services on Linux is entirely routine for us.

If there are many internal applications and they change constantly, an in-house specialist earns their keep: they understand the business from within. With that split we keep the hardware, the round-the-clock duty and the specialist knowledge, while they keep the applied work.

Describe your server room

Укажите количество машин, перечень работающих на них систем и того, кто их сейчас ведёт. Сделаем обследование и назовём первоочередные задачи.