As an estimated 92% of enterprises have adopted hybrid and multicloud strategies, according to the 2021 State of the Cloud Report from Flexera, cloud operations (CloudOps) teams face increasing pressure to simultaneously manage costs while improving business outcomes.
What levers can CloudOps teams pull to achieve operational objectives such as reducing hybrid and distributed cloud complexity, enhancing security, and automating processes? I put this question to cloud leaders from Hitachi Vantara and AWS in a virtual webinar last week. We convened during the week of AWS re:Invent to discuss the promises and pitfalls of CloudOps.
For example, one of the significant challenges CloudOps teams face in the hybrid world is building expertise for managing multiple types of cloud platforms, said Robert Hallahan, Sr. Specialist Solutions Architect at AWS. To meet this objective, Hallahan said CloudOps teams should practice the tenet of “you build it, you run it, you own it.” So gone are days when development teams can throw workloads over the fence for IT teams to operate and manage it – This essentially points to “necessity” around breaking down the walls between DevOps and ITOps teams with “Integrated Cloudops” approach that brings these two worlds together.
Good cloud work is iterative, Hallahan contended. Consequently, CloudOps teams should constantly look for opportunities to improve. He recommended focusing on understanding KPIs and service level objectives while iterating on applications, workloads, and operations to make and measure improvements.
Krishnaprasath Hari, Vice President and Head of Cloud Modernization at Hitachi Vantara, agreed and added that managing cost cannot be an afterthought. CloudOps teams that bring the right level of visibility, control, and operations expertise ensure “the variable spend” of the cloud is well-managed, “so cloud resources don’t become a sprawl.”
The Top 3 Steps to Improve Your CloudOps Chops
After this great discussion, which you can watch here, I put Robert and Krishnaprasath on the spot and asked for the top three priorities for CloudOps in the coming year. Here’s what they had to say:
Robert Hallahan, AWS:
1. Iteration of base frameworks. Frameworks aren’t something you can “just improve upon today and tomorrow, but always looking forward,” said Hallahan. To improve governance, he recommended “touchless deployments and touchless commits, from starting your pipeline to the end — being able to roll back code automatically.” These steps enable automating remediation, which he said is “one of the really big next steps we’re going to start seeing soon” for CloudOps teams.
2. Observability of your cloud platform. As customers look to migrate and modernize what he terms monolithic-type workloads, “you need to rethink how you collect output signals,” such as logs, metrics, and traces to improve the observability and evaluate the health of their cloud application.
3. Attribute KPIs to costs. Examine your usage and provisioned resources, said Hallahan. There are two key questions to consider. First, “Do they drive and lead to business outcomes?” And second, “Are there places in which we can leverage this type of data we’re collecting from our workloads?” Once CloudOps teams identify their vision of long-term success, he recommends they ask, “what are the things we can do to get there today?
Krishnaprasath Hari, Hitachi Vantara:
1. Reliability. Design with resiliency. Your cloud platform should be “always-on, autoscaling, and it can provide the right level of agility required — so that’s the foundation,” he said.
2. Autonomous. Make your cloud operations more autonomous. With 360-degree observability, Hari said you can ensure you’re tracking the right signals and making sure that “noise is illuminated.” He says that it’s critical to achieve “touchless operations,” which entails automating infrastructure and application management.
3. Outcomes. Translate operations into business outcomes. Examples may include improving KPIs such as revenue per transaction, or reducing cost per transaction, he said. “These are some of the practical ways in which your cloud operations can evolve and support your cloud-native workloads.”
As we prepare for the new year and all that it may bring, CloudOps will play an increasing role in the shift to hybrid and distributed cloud environments, resulting in “touchless” frameworks of infrastructure, applications and data.
Samta Bansal is Global Consulting Strategy and Marketing Leader at Hitachi Vantara.