Traditional IT operations management is under assault. Business leaders are openly adopting SaaS offerings without consulting IT, and DevOps teams are bypassing traditional IT Ops to deploy directly to the public cloud. More and more, IT operations is on the defensive, with open calls for a new no operations (NoOps) model.
Of course, business and development leaders still expect IT operations management to make cloud-based services and traditional systems work together, and this is the strongest case against NoOps. NoOps doesn't work, and your organization should avoid it.
Here are seven reasons why.
1. Maintaining legacy services is not just about the services
The evolution from physical to virtual machines, and to private and public clouds, has been swift—so swift that decades of legacy workloads have been left on islands, unable to participate in all of the benefits that the cloud has to offer.
Yet these workloads underpin services that are being modernized while still depending on legacy infrastructure, data, and applications. For example, banks' and airlines' mobile apps depend on customer, account, and flight data that resides on mainframes. Maintaining the performance and availability of legacy systems and services remains critical to digital business transformation efforts.
New services are using microservice and container architectures. As this trend continues, data centers may shrink, but the important workloads running within the new architectures will be more concentrated. The business needs a skilled IT operations management team to support them.
2. Edge computing will drive more, not less, demand for IT support
Edge computing—where devices take on more processing of data and react both independently and in concert with other devices—will be the next disruptive technology for IT. Business units are experimenting with smart technologies that promise to deliver better services at lower costs. But they just assume that IT will support these devices and the data they produce.
What they are asking for is extended infrastructure management. For example, a hospital might want smart wheelchairs and beds to track patients and patient care. But once they are happy with the pilot program and start to scale up, they will assume that they can throw support over the wall to IT. But this requires resources that don't even exist today in many IT operations groups.
To respond, IT Ops will need to embed operations staff into the business units. This trend is already happening in organizations that are scaling out DevOps practices, and edge computing will accelerate it. The result will be more demand for IT operations, not less.
3. You want business intelligence? You better have an ops team
Data collection, storage, and analysis are specialties of IT operations, born from a need to measure and improve performance. But data is for the business, not just IT. As more businesses conduct production and sales digitally, demand for digital data collection increases.
That data collection process needs to happen across a wide variety of systems, from the mainframe to the cloud, to feed the machine-learning algorithms that will fuel digital business changes and growth. IT operations must support all processes for collecting, storing, analyzing, and distributing reports across the hybrid environment, or the business's investment in AI will be for nothing.
4. Regulatory compliance and security don’t happen on their own
You can’t outsource compliance. Regulations apply whether your app runs on-premises or in the cloud. Someone has to patch servers to close vulnerabilities, administer access to sensitive information, and keep an eye on the activities of privileged users.
Security teams often rely on IT operations to manage network policy, administer identity governance, and enforce controls so that they can focus on threats, incident response, and policy-writing. Eliminating the operations team would be offset by the need to grow the security team, which would need to perform those functions.
5. Operators are learning to code, but developers are not learning to administer
Yes, DevOps is blurring the lines between developers and operations staff, and that has a positive effect when the team that builds an app knows that it must also support it. That tends to reduce the demand for a daily reboot of a server, rather than a real fix for a memory leak. Many IT operations professionals who join DevOps teams are now learning to code so they can provide automation to support the pace of DevOps.
But you will find very little information at industry conferences on how to identify and reduce the costs and risks of overprovisioned and unmanaged IaaS instances, or on how to architect the right level of redundancy to achieve the availability goals set by the business for a service. Nor will you find information about how to budget for future expansion of the network to support more employees as the business grows. There will continue to be a demand for these skills, independent of DevOps practices.
6. Incident and problem management requires a breadth of skills
When something goes wrong in a large environment, isolating and recovering from the problem can be a challenge for even the best teams. While bad code causes some problems, other issues may be related to network, security, storage, or capacity problems. Having a breadth of skills when incidents occur is critical to minimizing outages.
Also, teams that provide first response to incidents often need to be available 24x7. Developers may be used to being part of an escalation that sometimes happens at odd hours, but dealing with constant, round-the-clock incidents is an entirely different discipline. You must build an organization that's designed to handle it.
7. The cloud isn’t foolproof
Even the cloud pros can take hours to recover from a major outage. The idea that the staff and technology at a cloud service provider (CSP) can handle all of the operational needs of a business is laughable—the CSP is only responsible for its services, not your business’s digital strategy. And when the cloud is the disaster, who will decide to invoke the disaster recovery plan? Experienced IT operations professionals are in the best position to provide DR planning, training, and execution.
Why do some favor NoOps?
Web-scale companies, which lack decades of accumulated IT infrastructure and processes, have pioneered approaches that rely heavily on public cloud infrastructure and tools. The technologies have evolved rapidly, and provide an alternative to IT Ops-provided infrastructure.
As a result, we now have containers filled with microservices running in serverless architectures. Developers are more comfortable deploying their own code, and with rolling it back if necessary in a “fail fast” approach, rather than waiting on a cumbersome change management process for approval. Servers are no longer pets, but cattle—if one fails you just replace it, rather than nursing it back to health.
Meanwhile, business users are clamoring for SaaS, not because those services are delivered from the cloud, but because they’re better than what IT teams have provided, and because business stakeholders can acquire them easily. But if the business contracts for these services independently of the IT organization, then what role should IT operations play in managing them?
With the business and development teams demonstrating a strong preference for NoOps, should IT operations shrink to support only legacy systems that have yet to be retired? What possible future would an IT operations organization with decreasing responsibility have?
Hybrid IT management is what delivers
The seven arguments against NoOps are by no means comprehensive, but they hightlight the roles that IT operations must continue to play in the delivery of IT services. Roles and skill requirements will change with advancements in technology, as has always been the case. And where IT operations organizations fail to keep up—for example, in modifying change management processes to keep pace with the speed of DevOps—alternatives should be pursued.
And embedding IT operations into DevOps teams, or even directly into business units, is worth considering. But maintaining an effective IT operations team is in the best interests of businesses that want to support new digital business initiatives while also sustaining the services on which those businesses have depended to get to where they are today.
NoOps is nothing more than a bad idea.
Keep learning
Choose the right ESM tool for your needs. Get up to speed with the our Buyer's Guide to Enterprise Service Management Tools
What will the next generation of enterprise service management tools look like? TechBeacon's Guide to Optimizing Enterprise Service Management offers the insights.
Discover more about IT Operations Monitoring with TechBeacon's Guide.
What's the best way to get your robotic process automation project off the ground? Find out how to choose the right tools—and the right project.
Ready to advance up the IT career ladder? TechBeacon's Careers Topic Center provides expert advice you need to prepare for your next move.
