The term “auto-mation” comes from “(ε)αυτός“, the Greek word for “self“. Automating the IT infrastructure means making it capable of fulfilling its purpose without human involvement.

Few years back Gartner introduced three IT automation tiers: IT tasks, IT services and IT processes. Gartner implies a linear progress from tasks to services and finally to processes. We will now try to amend Gartner’s classification towards a continuous improvement model. In real life, IT evolves continuously and so does automation:

• Tier #1 – Automated IT Infrastructure Tasks. Automation targets specific tasks on infrastructure components, such as network, server, and storage or cloud platform components. Automation is implemented with scripts or runbooks, and is siloed across components. Examples are scripts to create VMs, assign IP addresses, add IP address to DNS etc.
• Tier #2 – Automated IT Infrastructure Services. Automation targets the coordination (orchestration) of multiple tasks to manage IT infrastructure services consisting of multiple components. Examples are scripts talking to multiple components to bring or operate an IT service, such as enlarging a DB, which usually includes multiple tasks such as adding storage, reconfiguring the DB, giving access to users, configuring backup etc.
• Tier #3 – Automated IT Infrastructure Processes. Processes refer to consumers of IT infrastructure, as described in ITL. These are not necessarily people in IT, and may have no knowledge of what is happening inside IT for a specific business function they are getting. Process automation refers to the flow of actions required to fulfill a business purpose. For example, adding a new business partner means that network access should be configured, partner number assigned, applications and databases updated etc. Automation in this tier means orchestrating these activities, some of which may not related to IT at all, in one process flow.

Automation Maturity

There is no standard method of measuring the automation maturity in IT operations. We can find an analogue in cars and follow the Autonomous Drive classification (see here), in which 5 automation maturity levels are defined:

Level	Name	Description
0	No automation	Entirely manual operations
1	Sporadic uncontrolled scripts	Self-motivated system administrators write scripts to facilitate their work. No standards, no framework, escalating efforts to cope with unmanageable scripts.
2	Assisted Manual Operations	Operations are still manual, but operators are assisted by scripts and advanced tools for repetitive or complex infrastructure activities. The scripts are properly documented and maintained in a life-cycle management framework.
3	No direct human interaction	Operators do not manage the infrastructure directly. They don’t need to login to components (such as servers). All activities are scripted. There are IaC tools (Infrastructure as Code), which manage the configuration of all components.
4	System is controlled by automation	The infrastructure is fully virtualized and abstracted (see this post) and is operated by automation. Operators work only proactively. They may communicate with automation at a higher level, putting requests for the infrastructure. How, even if, the requests will be honored is decided by automation according to predefined laws and rules.
5	Fully Autonomous Operations	Human operators are not needed to operate the infrastructure! All activities related to the infrastructure are decided and executed by automation. Any modification to the infrastructure behavior is described with laws and rules.

Using these automation levels, it is possible to asses the automation maturity for a specific IT environment. In the end, we may use a ‘radar’ diagram describing the overall IT operations maturity for automation, as in the diagram below.