The good and bad of ‘zero-touch’ cloud operations

Zero Touch Cloud Operations is the concept of automating cloud operations processes (cloudops) to minimize the need for human intervention. Specifically, it involves automating deployment, configuration, scaling, monitoring, and troubleshooting across all cloud platforms (public and private), and even extends to traditional legacy systems and edge computing systems as well.

Physically, it is a layer of technology that runs on top of cloud providers (for example, a metacloud). It can be anything that can automate cloud operations, for example, AIops technology, identity management, performance management, governance, and finops. This includes detecting problems and solving them without human involvement.

Adopting a metacloud also means less urgent operational processes, such as standard maintenance practices (including those related to disaster recovery). Devops toolchains also use these types of technologies for the same purposes.

Like any good tech move, there are pros and cons to consider. Let’s see both.

Good news

Basically, contactless cloud operations provide some well-understood advantages:

• Greater efficiency. Automating routine tasks in cloud operations frees up time and resources for cloud teams to focus on more strategic initiatives.
• Less bug. Most of the time something goes wrong, a human has messed it up. Zero touch reduces the risk of human error.
• Faster time to implementation. Automating deployments and scaling cloud services compresses the time frame from idea to implementation.
• Improved uptime. By using AI-based algorithms to predict problems before they occur (as happens in AIops), cloud teams can proactively resolve issues and improve uptime.
• Better collaboration. Integrating DevOps practices into cloud operations can improve collaboration between development and operations teams, with all the obvious benefits.

Bad news

Of course, the good news must be balanced against the realities that few discuss these days.

• Lack of control. Automating routine cloud management tasks can reduce the level of control cloud operations teams have over the environment. Many cloud architects are afraid of automation, but they have some valid points.
• Loss of cloudops abilities. If we don’t directly control the clouds most of the time, how will we learn to fix things when humans are needed? Zero-touch cloudops could lead to complacency. This could lead to disaster when real human skills are needed, but no one has done any real work in clouds for years.
• Balance automation with human supervision. Most companies err on one extreme or the other: either humans are never involved or they are fully involved. The reality is that zero contact still requires human supervision, but a balance must be found.
• Technology dependency. The double-edged sword is that technology can improve efficiency and reduce risk, but it can also introduce new risks and challenges if it fails. Many teams only focus on benefits. Some risk still exists, or net new risk may be introduced by contactless cloudops.
• It is expensive. Zero-contact implementation comes at a high price, not just the technology, but also the talent and planning to pull it off successfully. I often see companies trying to do it on the cheap and making things worse. If you think the cost savings you could result will fund a no-contact transformation project (I hear it often), you will be disappointed.

Is it worth the time, risk and money?

I have mainly focused on the downsides of zero touch here. Many other experts and technology providers will not. I am trying to bring all the facts to the table for your consideration. For most companies, investing in touchless cloud operations, including building a robust layer of metacloud technology, will bring a great deal of value to the business.

However, success is highly dependent on a specific domain, in terms of what it does and the solutions it creates. There are no contactless canned solutions yet, although many vendors claim theirs do in their sales guarantee. It requires planning and a lot of smart and expensive people to find the right solution. I suggest you make the necessary investment or don’t bother.

Hopefully I’ve lowered their expectations a bit, but I’ve also increased their chances of success with zero contact.