Choosing the Right Cloud Platform in South Africa
There is no universally best cloud. There is a best fit for a defined workload, organisation and risk boundary, and in South Africa that fit is shaped by local regions that do not all offer the same services, a bill priced in dollars and paid in rand, recovery geography, connectivity, power and scarce skills. This six-part series sets out a decision method that starts with the workload, applies hard gates before any weighted score, and refuses to treat an unknown as a pass. It then looks honestly at when Azure, AWS, Google Cloud, Huawei Cloud, Oracle Cloud and South African providers should win, when on-premises and private cloud remain valid, and when hybrid and multicloud earn their cost. One fictional South African group carries the examples, and Part 1 includes an interactive decision workbook. Not legal advice: regulated workloads need their own legal and sector review.
6 articles
- Part 1Choose the workload before the cloud: a South African decision frameworkThere is no best cloud, only a best fit for a defined workload, organisation and risk boundary. Part 1 sets out a decision method for South African organisations: hard gates first, weighted scores second, and a decision statement someone can challenge.13 min read
- Part 2When Azure is the right answer, and when it is only the familiar oneAzure is usually strongest where an organisation already runs on Microsoft identity, data and tooling. Part 2 separates the integration that removes real work from the familiarity that only feels like a reason, and covers the South African recovery caveat most Azure designs gloss over.10 min read
- Part 3When AWS or Google Cloud should leadAWS and Google Cloud often share a shortlist but tend to win for different reasons: AWS through engineering maturity and managed-service breadth, Google through data, analytics and Kubernetes. Part 3 sets out when each should lead, what their on-premises options really are, and how to settle a disputed advantage with an evidence spike instead of a demo.10 min read
- Part 4Where Huawei Cloud, Oracle Cloud and South African providers fitA credible cloud strategy assesses more than the three largest hyperscalers. Huawei Cloud, Oracle Cloud Infrastructure and South African providers each win specific workloads, and each should be tested with the same evidence, not used as a bargaining chip.11 min read
- Part 5When on-premises and private cloud remain the right answerCloud first should mean cloud is considered first, not that every workload must leave the building. Part 5 sets out the five defensible reasons to keep a workload local, what on-premises really costs, and how to judge the public-cloud appliances that promise the best of both.10 min read
- Part 6Hybrid and multicloud without architecture theatreHybrid and multicloud are consequences of requirements, not goals. Part 6 sets out when each earns its cost, why multicloud does not remove lock-in, the four levels of portability, and a seven-question test to apply before adding a second strategic cloud.11 min read
Working through any of this in your own organisation?
These articles come out of delivery work rather than theory. If the problems in this series look familiar, a short scoping conversation is usually the quickest way to work out what is worth doing first.
Talk to CloudNala