VP of Worldwide Field Engineering – Cloud and Datacenter at Canonical, the publisher of Ubuntu.
It’s accepted as a given in the tech industry: As companies everywhere up their digital game, automating time-consuming manual processes has become critical for making IT infrastructure more scalable, adaptive and efficient.
Survey after survey shows that companies are investing heavily in intelligent automation technologies, believing it will make them more competitive. The Covid-19 pandemic has added urgency to automation initiatives as organizations look to address cost pressures, support remote workforces, and increase agility and resilience in their supply chains.
Day-to-day reality tells a surprising story, however. Despite automation’s clear benefits, manual tasks remain prevalent inside many organizations. Many businesses are increasingly adopting automation, but they’re falling short in realizing its full value due to concerns over cost, integration, security and compliance.
A recent commissioned study conducted by Forrester Consulting, on behalf of Canonical, has found that application deployment and delivery professionals “are still continuously overwhelmed with the number of manual tasks and volume of work.”
Only about one-third of tasks are currently automated across IT functions, according to the study, with security, compliance and application lifecycle management the biggest areas weighing down teams. Nearly half of large enterprises are fulfilling key tasks on a completely manual basis.
“Despite the impetus for organizations to automate application lifecycle operations, it is clear that most companies still have a lot of work to do to reduce the number of manual tasks, especially the very large enterprises,” the study said.
What is happening here? If automation is so attractive for managing infrastructure and networks that are growing in size and complexity, whether on-premises or in the cloud, why are so many companies clinging to old manual methods?
Our work with customers shows that implementation cost and operational complexity is one of the biggest worries of companies looking to institute application development and infrastructure automation, especially when they’re planning to scale beyond the data center and cloud to the edge.
In addition to cost, they’re often concerned about difficulties in integrating different automation tools used by infrastructure and development teams.
It’s like the old quote: “Worrying is like sitting in a rocking chair. It gives you something to do but doesn’t get you anywhere.” Enterprises need to get to the automated future quickly because the costs of failing to do so are high. Let me illustrate.
Imagine a company — say, an international logistics firm — has invested heavily in digital technology to manage its transportation fleet, interact with customers, schedule deliveries, etc. Despite that, however, it still maintains an archaic manual process in one of the most basic aspects of application deployment.
Every time a developer wants to put a new application into production, they fill out a form asking the IT operations team for the required infrastructure — this number of virtual machines, that amount of storage, etc. Maybe a few weeks later — or if they’re lucky, a few days — the developer receives an email that says, in a nutshell, “Here are your servers and the IP addresses, you can log in with your credentials and they’re yours.”
Think about the waste — the human labor for such a simple transaction and the time wasted in getting an innovative app into customers’ hands.
The company’s competitor, however, has automated this routine process so that intelligent code in the application itself is able to send a request to the infrastructure and provision what it needs by itself. No human is involved; it all happens in a snap. For this company, deploying applications is faster, cheaper and more efficient.
Another area where manual processes can gum up the works involves how applications talk to each other, how they integrate with each other.
The old way is for applications to be initially configured by a human or by scripts written by a human and maintained by the same human. A newer method that got introduced was configuration management, whereby tools are used to automatically create configuration scripts. That method is better than purely manual, but there’s a problem: If something needs to change during the operation of that application, a person is still needed to intervene and create a specific automation script.
The most recent method of automated operations, however, involves creating an operational bubble around each application (some call it “operators,” “ops code” or “glue code”) that helps each app integrate with another automatically. Intelligent code tells the app how to deploy itself in different places within the infrastructure, how to scale while keeping all its integrations intact, how to automatically upgrade itself without downtime and so on.
As these illustrations show, automation is simply indispensable in operating infrastructure and applications at the speed and efficiency that the digital world demands. While it’s understandable why many enterprises have concerns about various aspects of the automation journey, to stick with manual processes is to deny the reality that the future of enterprises is automated.
So what steps should organizations take to overcome the operational complexity that can come with implementing automation, and how can they integrate automation tools as seamlessly as possible?
The answer is to leverage cloud technologies at the infrastructure and application levels. Enterprises should embrace tools like Kubernetes, an open-source system that helps automate the process of deploying and operating the applications that their developer teams pump out.
Put another way, if an organization finds itself struggling with a bevy of manual IT processes mostly because it’s been cloud-shy and is operating its own infrastructure, it should stop operating its own infrastructure.
By adopting a multicloud model — a combination of public and private cloud with a third party hired to manage the private cloud — companies can best leverage automation for predictable operational costs and maximum efficiency. A multicloud architecture can make the automation journey easier by allowing for wholesale reuse of prepackaged automation components.
A worthwhile goal for organizations is 100% IT automation, especially in infrastructure and application operations. The more automation a company can implement, the more it can free up IT staff and developers to focus on digital innovation that benefits customers.