What "Managed" Actually Means in Managed Cloud

Managed" appears on almost every hosting page, and it can mean anything from a monthly operating system patch to a team of engineers who run your whole platform. Two providers can both say managed cloud and be selling very different things.

Before comparing prices, find out three things: which layers the provider actually operates, what happens when something breaks, and how much of its engineers' time you are buying. This article walks through each one. For a short definition, see the glossary entry on managed cloud. For how responsibility shifts across hosting types, see which parts of security are yours and which are your host's.

What managed cloud is, and where private fits

Managed cloud is cloud infrastructure that a provider's engineers operate on your behalf. They keep it patched, watch it, fix it when it breaks, and help you change it. You still get the cloud parts: virtual machines you can create and resize, networks you can define, and storage you can attach. The difference is who keeps all of it running.

Whether the environment is private is a separate question.

Private means your environment is yours alone. Your workloads do not share compute or storage pools with other customers' workloads.

Managed means someone else's engineers run it day to day.

You can have either without the other. Public cloud is usually neither private nor managed by default. A self-run private cloud is private but not managed. The offers worth comparing are the ones that are both, and the question is what "managed" covers in each.

The layers, and who runs each

Every hosting environment is a stack. "Managed" means the provider operates some of those layers for you, and the honest question is where the line sits.

Physical infrastructure and hardware. Data centre, power, cooling, servers, disks and their replacement.

Hypervisor and cloud platform. The software that turns hardware into virtual machines, networks and volumes. On InMotion Cloud this is OpenStack.

Network and storage. Routing, firewalls at the platform edge, storage clusters and their health.

Guest operating systems. Patching, hardening and monitoring the operating system inside each virtual machine.

Middleware and databases. Web servers, database engines, caches and the services your application sits on.

Your application, data and access. Your code, what it stores, and who can log in to it.

Most offers draw the line at one of three points. Infrastructure-only management stops below the guest operating system. Operating system management adds patching inside your virtual machines. Fuller management reaches into middleware and databases. Almost nobody manages your application or decides who gets access to your data, and you would not want them to.

Infrastructure stack showing the three common points where managed hosting stops, with the application, data and access layer always remaining the customer’s responsibility. — Infrastructure, operating system and fuller management stop at different layers. Confirm the boundary in writing.

Our pricing page shows how InMotion Cloud plans scale by managed engineer hours rather than by locked features.

Five questions that tell you what "managed" means here

Which layers do you operate, and is that written down? A diagram in a sales deck is not a contract. Ask for the scope in the service terms.

Who patches the guest operating system, and how quickly? This is where offers differ most. Some patch on a schedule, some only on request, some not at all.

What happens at 3 a.m.? Ask who is watching, what triggers a response, and whether the first person you reach can fix the problem or only log it.

Are engineer hours capped, and what happens past the cap? Managed services are delivered by people, so the time is always finite. Find out whether extra work is billed, queued or refused.

What stays with you? A good provider will tell you plainly, without being asked twice.

Comparison of a tiered support queue with several handoffs before an engineer and an engineer-first model where the first contact can fix the problem. — Engineer-first support reduces the handoffs between raising a request and reaching someone who can resolve it.

What managed doesn't mean

Your application is still yours. The provider keeps the platform running. It does not write, debug or release your code.

Your data decisions are still yours. What you store, how long you keep it, and who can see it.

Access is still yours. The provider can enforce controls you choose, but deciding who gets in is your call.

Your compliance obligations are still yours. A provider can carry part of the work and give you evidence for its part. It cannot be compliant on your behalf.

It is not unlimited hands. Managed hours are a budget. Spend them on the operations work you least want to do yourself.

If the question you are really asking is how much work a managed platform takes off your team, our post on what managed cloud takes off a small IT team covers that in detail.

How InMotion Cloud defines managed

InMotion Cloud has two products, and both are managed and private. Virtual Private Cloud comes in four fixed plans, from Explorer to Large. Managed Private Cloud is built to your requirements for larger or more complex environments.

On both, InMotion Cloud manages the infrastructure stack: hardware provisioning, hypervisor and operating system patching, network configuration, security monitoring and storage management. Requests go straight to senior engineers who know your environment, with no first-line queue.

Every plan includes the same managed services: monitoring, backups, security patching and hardening, certificates, managed databases, load balancing, incident response, migration support, cost optimisation and advisory. What changes between plans is how many engineer hours each month they include:

Table listing the draft’s Explorer, Small, Medium, Large and Virtual Private Data Center plans with monthly prices and included managed engineer hours. — Compare monthly pricing alongside the engineer time included in each plan.

What stays with you is your application, the data it holds and the decisions about who can access it. We would rather say that here than have you find it in the small print.

Bar chart of managed engineer hours included per month on each InMotion Cloud plan in the supplied draft, rising from 3 hours on Explorer to 40 hours on Large. — The supplied plan comparison scales included engineer time from 3 to 40 hours per month.

Summary

"Managed" is a scope, not a promise. Find out which layers the provider operates, how incidents are handled, how many engineer hours you get and what happens past them, and what stays with you. A provider that answers all five plainly is usually the one worth talking to.

Want to know exactly what we would manage in your environment?

InMotion Cloud engineers will map your current setup layer by layer and show you where our work ends and yours begins.

Talk to a cloud architect

Frequently asked questions

What is managed cloud? Cloud infrastructure operated on your behalf by a provider's engineers, who keep it patched, monitored and running while you use it.

What is the difference between managed and unmanaged cloud? With unmanaged cloud you get the infrastructure and run everything above the hardware yourself. With managed cloud, the provider's engineers operate some or all of the stack.

Is managed cloud the same as private cloud? No. Private describes who the environment belongs to. Managed describes who operates it. An offer can be one, both or neither.

Does managed cloud include application support? Usually not. Providers manage the infrastructure and often the operating system and databases, but your application code stays your responsibility.

How are managed services priced? Most providers include a set amount of engineer time or a defined scope in the monthly price. Ask what happens when you need more than that.