With the increasing popularity and adoption of native cloud and serverless systems organizations are starting to demand that DevOps better understand – and manage – cloud costs. This shouldn’t come as a surprise. Although the cloud promises many things, including lower costs due to not having to buy and manage the infrastructure, in reality, organizations who have committed to the cloud are actually seeing cloud costs increase – sometimes in a dramatic way, and are often surprised when the bill arrives. As a result, executives and others in the organization are starting to ask some hard questions of development and operations:
- Why I am I always surprised when the bill arrives?
- How can I better plan and budget for total cloud costs?
- How do we better manage our reserved instances?
- Where do we begin to optimize our cloud deployment?
And more important questions that affect the bottom line that need to be understood –
- How much does this product or service cost?
- How profitable is this new feature?
- What is our true cost of goods sold?
Today these questions may seem hard to answer. At best, Development or Operations may have access to reams of log data about system performance but they find it hard to interpret costs. The “needle in the haystack” – cost does not rise out of the reams of data as an important operating metric. There is no single trusted source of data that enables Finance, Operations and others to speak the same language to effectively manage costs.
This burden typically falls upon Operations or Finance, or a combination of the two – for good reason. Finance is charged with managing the cloud bill and ideally can lean on Operations as a skilled and multilingual organization that understands how to speak to both Finance and Development.
What’s needed in a perfect world?
The ability to understand operational cost at an architectural level, repeatedly and in real time, enabling Finance, Operations and Development to understand, interpret and contextualize cost related information across products, teams, departments, and the organization as a whole.
Most of today’s cloud management tools were not built to be cloud cost management tools and were not designed to observe granular, real-time operational cost.
Finance, Development and Operations working together in the perfect world: FinDevOps
In a perfect world, Finance, Development and Operations would have an automated and real-time cost review of their cloud systems, continuously showing developers the cost ramifications (good and bad) of their design decisions. The architecture would be visualized and overlaid with cost and context, enabling a cloud system to be cost-architected in addition to being well-architected.
To do this the organization would embrace Finance in the DevOps culture. FinDevOps would establish cost-based collaboration across the enterprise, teams and disciplines. This perfect world is one where organizations leverage cost as a true, first-class operational metric.
- Imagine the ability to understand costs in real-time – to be able to understand how much a new micro-service deployment is costing the enterprise, the moment it is deployed.
- Imagine a world where understanding cost of goods sold is presented visually, to the executive team in the same way Operations evaluates site traffic or latency.
- Imagine Operations seeing cost as an ebb and flow across the enterprise and understanding the relationships of how a new deployment, traffic spike or unpopular product might impact cloud costs – in real time.
- Imagine knowing beforehand if an anomalous spike might result in a devastating increase to your cloud bill.
- Imagine having the means to predict and identify cloud cost trends in the most granular way.
- Imagine teams, departments and the organization all working together focusing on a single source of truth of cloud costs.
And finally, imagine being able to achieve all of this without spending months instrumenting agents, integrating API’s and training?
Now imagine CloudZero. Delivering on the promise of FinDevOps. Leveraging cost as an operating metric – across the enterprise.