Azure Myth 6: Cloud is expensive - Azure MythBuster

Azure Myth 6: Cloud is expensive - Azure MythBuster

2019-09-04

The perception that cloud is expensive often stems from comparing on-premises hardware purchase costs against cloud operational spend without accounting for the hidden costs of on-premises: hardware depreciation cycles, power, cooling, physical access control, and network connectivity circuits. A more accurate comparison treats cloud as a leasing model where you pay for what you use and can scale to zero during off-peak hours — something that rarely happens in on-premises environments.

Cost design starts with defining your cost envelope: a fixed budget (e.g., £5,000/month), a variable rate (e.g., cost per visitor or per transaction), or both. This envelope should be treated as a requirement alongside the five pillars of software quality — effectively making cost a sixth pillar. Auto-scaling is central to cost efficiency: VM Scale Sets, App Service, and Azure Functions can scale down to a minimal footprint during quiet periods and scale up automatically before predictable spikes such as TV advertising windows or Black Friday. Choosing the wrong abstraction layer — IaaS when serverless would suffice, or serverless when a fixed-price App Service plan is more predictable — leads directly to the wrong cost envelope. PaaS and serverless options also simplify cost forecasting because spend scales linearly with load, unlike the large step-function capex of on-premises hardware refresh cycles.

Related Content

Azure Myth 4: Azure is Magical! Management in the cloud compared with on-premises - Azure MythBuster

Azure Myth 4: Azure is Magical! Management in the cloud compared with on-premises - Azure MythBuster

2019-08-21 Microsoft

Moving workloads to Azure does not eliminate management decisions — scalability, resilience, and high availability all require deliberate configuration. This Azure Mythbusters episode contrasts scale-out via VM Scale Sets and auto-scale rules with scale-up by increasing VM SKU size, explains availability sets and availability zones, and shows how PaaS services like Azure Functions still require you to choose the right plan and design cross-region resilience with Traffic Manager.

Azure Myth 3: You don’t need requirements in the Cloud… Or do you? - Azure MythBusters

Azure Myth 3: You don’t need requirements in the Cloud… Or do you? - Azure MythBusters

2019-08-20 Microsoft

Requirements remain essential when migrating to or building on Azure — from availability SLAs and RPO/RTO targets to compliance, data sovereignty, and cost. This Azure Mythbusters episode uses composite SLA calculations and Azure Architecture Center reference architectures to show how under- or over-specifying requirements directly shapes your design, region strategy, and overall cost.

Deploying a multi-region Serverless API Layer (Part 1)

2019-07-13 · 4 min

In my spare time, I work on a pet project called Theatreers. The aim of this is a microservice based platform focused on Theatre / Musical Theatre (bringing a few of my passion areas together). I've recently re-architected the project to align to a multi-region serverless technology stack.