Managed infrastructure services: the layer under the workplace
Servers, storage, network, virtualisation and the cloud environment around them. Managed infrastructure services are the management of that layer: monitoring, maintenance, capacity and recovery, under one agreement instead of four separate ones.
What sits under it
The line runs at the workplace: what is above it is workplace management, what is below it is this.
- Servers and virtualisationPhysical and virtual servers, the platform underneath and the patching on top. Including when something has to be replaced, because a server running out of support is a risk with a date on it.
- Storage and backupCapacity, growth and the copies. Backup belongs here because it is the most forgotten and the quickest to fail quietly; a job that fails with nobody watching fails for months.
- Network and connectivitySwitches, firewalls, wifi and the lines between them, monitored on what actually passes through. That is the layer where an outage announces itself as "the internet is slow".
- Cloud and hybridThe environment at a cloud provider counts as infrastructure: rights, network separation, cost and capacity. Otherwise a second management practice grows beside the first, with a boundary nobody watches.
How we take it over
Three steps. The first takes longest and yields most.
Record what is there
A current list of systems, integrations, licences and end dates. In almost every takeover this is the moment things surface that nobody remembered, a server without an owner, a certificate expiring next month.
Switch on monitoring and maintenance
Monitoring that says not only that something is on but whether it works, and a maintenance rhythm that is fixed. Alerts nobody reads get switched off; an alarm that always sounds is not an alarm.
Report on what it buys you
Availability, what was resolved, what replacement is coming and where capacity is tight. Without that, management stays a cost line without a face, exactly what gets cut at the next round of savings.
What you keep yourself
Taking over the management is not the same as taking over the say. Four things stay yours, and that is deliberate.
- The applications and the dataWe manage the layer they run on, not what happens inside them. What an application should do and who may see which data is a decision for the organisation, not a setting for the administrator.
- Who gets access to whatWe set the rights up and make them visible, but you decide who gets them. Management that grants itself access is the first thing an audit points at.
- The investment decisionWe say when something runs out of support, on what date, and what the risk is if it stays. Whether and when it is replaced is your call against the rest of the budget.
- The building workPower, cooling, floor and walls of the server room are construction work and we do not do it. We lay the data cabling, mount the equipment and make sure the path out stays up.
How it is recorded
Management you cannot account for is management on trust alone. Four things are written down before they are needed.
- The list stays rightThe inventory from step one is not a one-off document but part of the work. A list that ages is worse than no list, because decisions get taken on it anyway.
- What was agreedWhat the management covers, which response time applies and what availability is promised. That belongs on paper before the first outage, not during it.
- What was changedEvery change with a date and a reason. Without that trail, a fault starting a day after an adjustment cannot be traced to it, and gets solved twice.
- What went wrongWhat the outage was, how the recovery ran and what changed afterwards. An incident that is only resolved comes back; an incident that is recorded changes something.
Further reading
What this connects to.
Questions we get about this
The ones that come up most, answered briefly.
Where is the line with workplace management?
At the workplace itself. What sits above it (laptop, image, support for the user) is workplace management. What sits below it, from server to network to cloud environment, falls under managed infrastructure services.
Do you take management over from another party?
Yes, and that always starts with recording what is there: systems, integrations, licences and end dates. In almost every takeover things surface that nobody remembered, which is exactly why that step is not skipped.
Do we keep control over our own applications and data?
Yes. We manage the layer they run on, not what happens inside it, what an application should do and who can see which data stays a decision for the organisation. Who gets which access rights is also your call; management that grants itself access is the first thing an audit points to.
Is the response time agreed before or only after the first incident?
Before. What falls under the management, which response time applies and which availability has been promised is set down before the first incident happens, not during. Every change and every incident is added with a date and a reason, so a fault that starts shortly after a change stays traceable.
Management you can account for
Tell us what runs and where it sits now. We say what we take over, what you keep, and where the handover carries risk.
Practical IT knowledge in your inbox
New guides on management, security and the workplace, written by the people doing the work. No sales talk, and you can unsubscribe in one click.