Cloud Concepts
What cloud computing is
Cloud computing is the on-demand delivery of compute, storage, networking and database capacity over a network with pay-for-use pricing. The NIST definition names five essential characteristics you should be able to recognize: on-demand self-service (no human ticket required), broad network access, resource pooling (multi-tenancy), rapid elasticity, and measured service (metering, which is what makes pay-per-use possible).
Service and deployment models
IaaS delivers raw infrastructure - virtual machines, block storage, virtual networks - and you remain responsible for the guest OS and everything above it. PaaS delivers a managed runtime or database, so you own data and configuration but not the OS or host. SaaS delivers a finished application. Deployment models are public, private, hybrid (on-premises connected to cloud) and multicloud (two or more cloud providers). A recurring trap is confusing hybrid with multicloud, and another is asking who patches the guest OS in a PaaS offering - the provider does.
Scaling, resilience, DR
Scalability is the capacity to grow; elasticity is automating growth to match demand and shrinking back to control cost. Vertical scaling resizes one machine; horizontal scaling adds instances. High availability survives component failure through redundancy across availability domains; disaster recovery restores service after a regional failure. DR planning uses RTO and RPO.
Rule of thumb
RTO = maximum acceptable downtime (how fast you must be back). RPO = maximum acceptable data loss (how much data you can afford to lose). If a scenario says "no data loss at all," the answer is synchronous replication with RPO near zero; if it says "cheapest, some loss acceptable," choose asynchronous replication and state the RPO in minutes.