In this five-part series, I’m taking a hard look at the common – and costly – mistakes organizations typically make while building a cloud architecture. Part one explained how organizations can quickly lose visibility and control over their data processing,and detailed how to avoid that mistake. Part two looked at why a DIY approach often goes wrong, and how an independent cloud networking platform solves that problem. In part three, I’ll examine how easily costs can mount when organizations don’t have a cloud networking platform that enables intelligent costing and billing.
After working with hundreds of enterprises and reviewing thousands of cloud network designs, I realized that most enterprises have poor cost-sharing models with various lines of business (LOBs). The root cause is the absence of an intelligent cost and billing solution that shows how other teams use shared and non-shared infrastructure resources. This lack of granular and flexible billing without a proper cost-sharing model often results in higher costs. Enterprises face issues such as:
- Unnecessary, underutilized, and duplicated shared resources
- Mergers/acquisitions halted or delayed resulting in expensive, disconnected, and siloed architecture
- Inaccurate, manual, and error-prone financial planning with off-the-mark resource utilization assumptions
Out of these issues, the utilization insights by LOBs and cost-sharing for shared resources are a significant concern. Examples of shared resources include private circuits (Direct Connect, Express Route, etc.), next-generation firewall (NGFW), storage, and so forth. Since proper showback and chargeback are absent, each team and LOB builds and orders services independently. In most cases, these services can easily be shared by various teams and LOBs to save costs.
Every CSP provides a billing dashboard, but those bills cannot itemize the shared resource utilization. Moreover, if a service is consumed in a multi-cloud scenario, there is no way that CSP1’s bill can show that LOB2 in CSP2 utilizes CSP1’s services.
Some FinOps products provide billing insights, but they are merely glorified CSP bills. Some go beyond that and offer cost-saving suggestions, which are still based on the bill generated by the respective CSP. What works best is if the cloud networking vendor – which sees the actual traffic and utilization – can visualize this cost directly.
Enterprises need a cloud networking platform to provide billing and utilization insights from the infrastructure perspective. This platform must be agnostic to CSP and CSP-specific billing. The telemetry must be part of the data plane so that actionable intelligence can be added to it.
This platform should allow you to define and identify apps and services belonging to LOB1 vs. LOB2 vs. central IT. With its robust cost-sharing model, shared resources must be utilized effectively among various teams with transparency to save a ton of cost.
Enterprises must invest in a cloud networking solution that allows them to provide deeper cost visibility using a rich set of parameters such as CSP tags, IP/CIDR, and regions, to name a few – and then allow the central IT team to bill it to respective departments and LOB. This approach ensures individual teams are not deploying their own LB, DX, storage, etc., and can effectively utilize the resource provisions.
Shared resources are at the core of cloud computing, but the lack of visibility into how those resources are used results in underutilized or duplicated resources, error-prone financial planning, and siloed operations that can also undercut mergers and acquisitions. Cloud networking that provides a comprehensive view of resources being used by LOBs or business units (BUs) and that enables a cost-sharing model can hold down costs.
Part four of this series will look at how an on-premise mindset inflates security costs while weakening defenses.