Web sites and apps
Managed hosting, TLS, managed databases.
What is cloud computing? Definition and service models
Cloud computing means accessing computing resources — servers, storage, databases, networks — on demand over the Internet without owning the physical hardware. A provider pools infrastructure and charges for usage. IaaS, PaaS and SaaS describe how far shared responsibility extends between you and the vendor.
In one sentence
Cloud is renting compute capacity by the minute instead of buying servers in a closet.
Key points
- Elasticity: scale capacity with load (campaigns, seasonality).
- Shared responsibility: the vendor secures the cloud; you secure what you deploy in it.
- Common deployments: public (AWS, Azure, GCP), private, hybrid.
- Cost can spiral without governance (forgotten instances, idle storage).
Term at a glance
- Cloud computing
- Infonuagique · Informatique en nuage
- Quebec French term
- Infonuagique
- Domain
- Cloud and DevOps
- Category
- Cloud
- Level
- Beginner to intermediate
What cloud computing really means
NIST's definition highlights on-demand self-service, broad network access, resource pooling, rapid elasticity and measured service. Spinning up a VM in minutes or off-site backups fit this model.
It does not mean everything must live with one hyperscaler. Many organizations pick regional or Canadian clouds for data residency and latency. The pattern is outsourced, operated capacity.
Lift-and-shift without redesign can cost more than a right-sized server. Cloud pays off when variable load or time-to-market outweigh recurring spend.
How organizations consume cloud
- 01
Pick a service model
IaaS (you run OS and app), PaaS (you deploy code), SaaS (you configure the product).
- 02
Provision with code or console
Infrastructure as Code (Terraform, Bicep) for repeatable environments.
- 03
Secure and segment
Private networks, identities, encryption, least privilege.
- 04
Watch cost and performance
Budgets, alerts, shut down test resources, pick region and storage tier.
A concrete example
A marketing agency runs a national contest — 20x traffic for two weeks. Instead of idle servers all year, it hosts on a PaaS with autoscaling and CDN. After the campaign it scales down to a small fixed plan. Spend follows revenue — elasticity for a mid-size business.
Common use cases
Backup and DR
Off-site copies, geo replication, restore tests.
Bursty processing
Queues, serverless functions, video rendering.
Analytics
Managed warehouses and BI without upfront storage buys.
Benefits and limits
- Fast start without hardware capex
- Scale with demand
- Managed services (patching, HA)
- Global regions and compliance offerings
- Vendor and connectivity dependency
- Complex billing without FinOps
- Egress fees and lock-in risk
- Sensitive data residency requirements
Cloud or on-premises?
| Cloud computing | On-premises infrastructure | |
|---|---|---|
| Upfront spend | Low capex, operational opex | Hardware and facilities |
| Elasticity | Minutes to hours | Weeks to procure |
| Responsibility | Shared with provider | Fully internal |
| Best fit | Variable load, fast growth | Stable load, total control |
Impact on your organization
Cloud is a procurement and operations model, not a goal. Leadership should ask about data residency, three-year total cost and internal ability to govern access. Used well it accelerates launches; poorly governed it becomes an opaque recurring line item.
Frequently asked
Is cloud automatically more secure?
No. Vendors invest heavily, but misconfiguration (public buckets, weak passwords) still exposes data. Security is shared work.
IaaS, PaaS or SaaS?
SaaS for standard needs (email, CRM). PaaS to run your code without OS ops. IaaS when you need full stack control or legacy migration.
Can we stay in Canada only?
Yes — major providers offer Canadian regions; local vendors too. Check contract and replication.
Related terms
Sources and references
Comparing cloud, dedicated hosting and on-prem for your next project?
Clarify your hosting