By E.G. Nadhan, Red Hat Global Chief Architect Leader, Field CTO Organization & Yar Savchenko, Comerica Bank SVP, Infrastructure
Hot off the presses from 2024 Innotech Dallas conference comes deep insights from varied perspectives based on experience working with different enterprises over the years, such as these:
- enterprises who were anti-cloud at that time
- who has been effectively operating in the cloud for some time now and
- those who are considering venturing into the cloud on a trial-basis
A Reality Check on Cost.
When new environments are provisioned or additional cloud services consumed, there needs to be an automated governance around why they are needed with cost justification, potential charge or show-back model and effective real-time monitoring of their consumption pattern. While developers are likely to go ahead and request new environments to be provisioned and/or new cloud services to be consumed, such governance is better best exercised through automated controls that are in place based on architectural guidelines. In addition, companies should consider utilizing cloud native scaling solutions (such as AWS ASG) to manage their provisioned resources based on usage in production. But, hold on … how about a Reality Check on Architecture!
A Reality Check on Architecture.
Like all solutions, cloud deployments need well-defined architectures to factor in the cost component. Pretty graphics with boxes and lines only go so far. These architectural guidelines are better enforced into the solutions deployed only through active engagement between the Architects and the Cloud Implementation teams.
A Reality Check on Resilience.
Resilience is not just about the service continuing to be available to the customer when there is a failure. The environment where the service failed must be preserved so that root cause analysis can be done to ensure that this does not recur.
So, what techniques can enterprises use to address these reality checks?
Chaos Engineering — the art of chaos engineering provides enterprises a fair opportunity to experience what happens when things go wrong in production (and they always do!). This is accomplished by injecting a common failure in a controlled manner, to test mitigation techniques, build engineering muscle memory and validate technology, people, and processes. Conducting chaos engineering exercises on a regular basis enables enterprises to design and validate their response to incidents as well as stress test their most critical systems to proactively identify architecture gaps.
Back to the three types of enterprises…here are the calls to action that the speakers have to offer:
- For those who have been in the cloud for some time – Have a good baseline for the metrics for the environment that used to be. This positions you to draw a fair contrast with the metrics for the cloud environment and answer the ever-prevalent cloud question: Are you getting value?
- For those who just ventured into the cloud – Reach out and ask others who have been there about the do’s and don’ts. What works? What doesn’t? Attending sessions like these would be a great start!
- For those who are thinking about going into the cloud — Have a plan! The ask to go to the cloud is coming around the corner!