Any engineering organization that has embraced the public cloud is looking for ways to be more productive. Even those with advanced cloud strategies know that manual and inefficient processes can distract their teams and stand in the way of the agility and flexibility benefits that made the cloud attractive in the first place.
Earlier this year, a panel of engineers participated in a discussion during our CloudLIVE Virtual Conference to share what their teams have done to navigate these challenges. Brent Strong, Manager of Cloud Engineering at Change Healthcare, referred to the importance of minimizing the “cognitive load” on developers—or the list of concerns that might distract from their top priorities—so that it’s easier for them to focus on what’s important.
Here are a few practical suggestions from the conversation that have helped engineering teams remove the barriers that prevent them from taking advantage of the benefits of the public cloud.
Automating start and stop times on instances
One sample measure that has made a difference is automating when instances will start and stop.
When creating applications or services that will not operate outside of certain hours, starting and stopping the instances to align with those hours will prevent paying for the infrastructure when it’s not needed. When engineers need to do this manually, they run the risk of forgetting and accidentally exceeding budget for a project. By tagging the instance to start and stop at the right times automatically, they reduce the risk of wasted spend and no longer need to worry about the implications.
Automating manual processes for Reserved Instance modifications
Many AWS customers take advantage of Convertible Reserved Instances (RIs) for the ability to change instance types while still taking advantage of discounts to On-Demand pricing. And while AWS Savings Plans offer a comparable mix of flexibility and discounts, it’s clear that many AWS customers still rely on RIs to maintain the coverage they need.
During the panel, Riley Jenkins, Senior SRE Architect at Domo, explained that Convertible RIs have long been valuable for the flexibility to adjust infrastructure as needed. In the early days, however, the process of changing instance types manually required a lot of time digging through spreadsheets, which took up valuable time that could be spent elsewhere. Brent from Change Healthcare agreed that automating the modification of RIs has made a tangible difference by eliminating tedious and unnecessary manual work.
Giving engineering teams the right data to make decisions about scalability
Riley also discussed how much time engineers can save when they have an understanding of scalability before moving forward with a project.
Specifically, Riley said his teams review data on cost, performance, and usage proactively so they can decide not only what it might cost to scale a project, but whether it will be repeatable or will meet performance needs. This enables the teams to understand scalability and make decisions accordingly, rather than spend time on projects before discovering that they will not scale as needed.
Creating trustworthy infrastructure as code
The decentralized nature of the public cloud can be an asset, allowing anyone to access the infrastructure they need, as they need it. However, this decentralization can also lead to inefficiency and security risks, as each team uses infrastructure in different ways without the oversight to ensure they’re using it correctly or most efficiently.
Establishing and sharing infrastructure as code can help strike a balance between oversight and developer productivity. On the panel, Brent highlighted the value of sharing templates of infrastructure that is configured to comply with the healthcare industry regulations that the organization needs to adhere to. This not only reduces the risk of misconfiguration due to human error, which could result in compliance violations and data breaches, it also eliminates the need for the engineering teams to recreate that infrastructure every time they need it.
Nathan Peck, Senior Developer Advocate at AWS, pointed out the importance of a Cloud Center of Excellence (otherwise known as a Cloud Business Office or Cloud Strategy Office) in establishing templates for infrastructure as code. Collaboration across those with insight into finance, security, and engineering will ensure that the templates for infrastructure are created with all implications taken into account.
You can watch the full panel discussion here. To learn more about how engineering contributes to a successful Cloud Center of Excellence, read our in-depth whitepaper, Benchmark Your Cloud Maturity: A Framework for Best Practices.
