IT automation pros and cons
IT automation’s benefits include faster data center and cloud operations; reduced errors and variation from one implementation of a task to the next; and enhanced security and governance. However, an IT automation strategy must account for and eliminate errors; an automated error will proliferate much more quickly than a manual error. IT automation can also erroneously become a goal in and of itself, regardless of the return on investment from the initial setup work to time saved.
IT operations requires a significant number of distinct tasks. An IT administrator can accomplish each task manually, but modern business demands place extraordinary pressure on IT staff to respond quickly to needs across large, complex infrastructures. Humans cannot provision and configure workloads in minutes and accomplish all of the individual routine tasks required, at any time of day. Automation frees administrators from time spent on routine tasks so that they can apply themselves to value-added projects for the business, such as IT infrastructure optimization and experimentation with promising new technologies and products.
While automation saves time, it requires that admins carefully plan and research each task necessary for the intended workflow and then correctly translate those steps into the automation platform to achieve the desired end state. A company may appoint one or more IT automation managers, replacing or supplementing the role of IT administrators.
An IT administrator is liable to make an error while typing in a CLI, choose the wrong configuration setting for a server, overlook a key step in a complex task or make other mistakes. Errors lead admins to take additional time to troubleshoot and repeat the work process to get it right. IT automation enables an IT professional to construct a proven, accurate sequence of operations that can be run countless times in exactly the same manner.
While countless repetition without deviation is a benefit of IT automation, it can also be detrimental. Errors and oversights are easily codified into an automated process, which the automation tool will perform as quickly and efficiently as it does the correct steps. If the administrator automates a complex sequence of events and misses a key step or sets a variable incorrectly, that error is repeated ad nauseam until it is caught, remediated and rolled back. The 2010 flash crash of the United States stock market damaged global trade because of an automated computer system with a flawed algorithm. Test and vetting procedures must be part of an IT automation strategy.
An automated system is not the same thing as an intelligent system; it only knows as much as the human that programmed it can distill into scripts and commands. For example, an email spam filter is an automated IT mechanism with the intent to filter out unwanted messages. Occasionally, valid email messages will end up in the spam folder, and unwanted spam email gets past the filter.
Different IT administrators perform the same task in different ways, and even the same administrator handles a task differently from one time to the next. For corporate governance and regulatory compliance, an IT automation strategy demonstrates consistency in IT operations, regardless of the administrator on any given day.
Processes change over time as the IT infrastructure grows and changes, and technologies and best practices evolve. Automated processes remain static until a person decides to change them. Organizations must have a set workflow to update and re-validate automation processes, including disciplined automation versioning that tracks how tasks change over time.
Integration and inter-operability
IT automation tools must be compatible with systems, software and other elements across potentially diverse IT environments. Ideally, an automation tool should integrate with higher-level orchestration tools to roll tasks together under governed workflows.
The future of IT automation
IT automation is hardly a new idea, but the technology is still in its formative stages. Even the most full-featured tools depend on an IT professional or team to develop and maintain discrete automation elements, such as scripts, templates, policies and workflows.
IT automation will progress to act with greater intelligence and autonomy. IT automation platforms are likely to rely heavily on artificial intelligence (AI) technologies. For example, an automation tool can synthesize data on configurations, performance and other information across an IT deployment and process these inputs to discover a normal system operations benchmark, a deviation from which would trigger corrective actions. IT automation systems imbued with AI insights theoretically lessen the importance of deliberate, human-made automation rules, relying instead on autonomous choices guided by high-level business cost and compliance requirements.
Process automation, robotic process automation and service automation
IT automation is a broad term, often conflated or bound together with business task automation. An automated IT workflow can accomplish a strictly IT task, such as provisioning additional storage to a VM, or a business task, such as creating a new user account on the corporate email system.
Process automation improves workflows, typically in factories and other settings, where the same task or series of tasks occurs repeatedly.