Platform as a service Study Guide

📖 Core Concepts Platform as a Service (PaaS) – A cloud model where the provider delivers a complete computing platform (runtime, middleware, OS, virtualization, servers, storage, networking) and the customer only writes and runs applications and manages its data. Primary goal – Remove the need for developers to worry about infrastructure and operations, letting them focus on code. Provider‑managed components – Runtime, middleware, operating system, virtualization layer, servers, storage, networking. Customer‑managed components – Applications (code) and the data those apps use. Deployment models Public PaaS – Hosted in a public cloud; provider supplies the full stack. Private PaaS – Runs behind a firewall, on‑premises or in a dedicated cloud environment. Hybrid PaaS – Mix of public and private resources to meet security/control needs. Position in the cloud stack – Sits between SaaS and IaaS: more control than SaaS, less than raw IaaS. 📌 Must Remember PaaS simplifies code writing by abstracting infrastructure and operations. Provider responsibilities: runtime, middleware, OS, virtualization, servers, storage, networking. Customer responsibilities: application code and data. Public PaaS = provider‑managed end‑to‑end; Private PaaS = customer‑controlled environment; Hybrid PaaS = blend of both. Advantages: higher‑level programming, built‑in auto‑scaling, easier maintenance. Disadvantages: higher cost at large scale, reduced operational control. All PaaS vendors offer application hosting, a deployment environment, and integrated services (e.g., collaboration, DB integration, monitoring). Language/framework support is a key differentiator; you write in a supported language, upload, and the platform runs it. Scalability and maintenance responsibilities vary by provider (vendor‑maintained vs customer‑maintained). 🔄 Key Processes Provisioning – Choose a PaaS provider, select a deployment model (public/private/hybrid). Instantiate environment – Provider spins up runtime, middleware, OS, and networking automatically. Develop application – Use supported languages/frameworks; design, code, test locally. Upload & deploy – Push code to the platform; platform handles containerization/VM placement. Scale – Platform auto‑adjusts resources up/down based on load (built‑in scaling). Monitor & maintain – Use provider’s instrumentation/monitoring tools; provider patches underlying stack, you update app code as needed. 🔍 Key Comparisons Public PaaS vs Private PaaS Public: hosted by provider, shared multi‑tenant, fast start‑up. Private: behind firewall, dedicated resources, higher security/control. Hybrid PaaS vs Pure Public/Private Hybrid: combines public scalability with private data protection. Pure: either fully public or fully private, no mix. PaaS vs IaaS PaaS: provider manages OS + middleware; you manage only app & data. IaaS: you manage OS, middleware, runtime, plus app & data. PaaS vs SaaS PaaS: you develop custom apps on the platform. SaaS: you consume a finished application; no development. ⚠️ Common Misunderstandings “PaaS = full OS control.” – OS is managed by the provider, not the customer. “PaaS is always cheaper than IaaS.” – At large scale, PaaS pricing can exceed IaaS due to managed‑service premiums. “Public PaaS cannot meet compliance.” – Many public PaaS offerings provide compliance certifications; the issue is data residency/control, not impossibility. “Hybrid PaaS = two separate platforms.” – It’s a single logical environment that routes workloads to public or private resources as needed. 🧠 Mental Models / Intuition “Managed kitchen” analogy – The provider supplies the stove, oven, and pantry (runtime, OS, storage). You only bring the recipe (application code) and ingredients (your data). Layered cake – IaaS = bottom layer (infrastructure), PaaS = middle layer (platform), SaaS = top frosting (complete app). 🚩 Exceptions & Edge Cases Cost explosion – When workloads demand massive compute/storage, the per‑unit managed‑service surcharge can make PaaS expensive. Operational control limits – Certain low‑level network or OS tuning may be unavailable on public PaaS. Vendor lock‑in – Proprietary services (e.g., specific databases or monitoring tools) can hinder migration. 📍 When to Use Which Public PaaS – Rapid prototyping, startups, and apps without strict data‑sovereignty requirements. Private PaaS – Regulated industries, sensitive data, or when you need custom OS/network configurations. Hybrid PaaS – When you need public‑scale elasticity but must keep specific workloads or data on‑premises. Choose a provider based on: supported language/framework, scalability limits, maintenance model (vendor‑ vs customer‑maintained), and compliance certifications. 👀 Patterns to Recognize Questions mentioning “auto‑scaling”, “managed runtime”, or “focus on code” → PaaS is the answer. Prompts that list “network, servers, OS, middleware” as provider responsibilities → identify the PaaS model. Scenarios where only the application and data are under the customer's control → PaaS, not IaaS. 🗂️ Exam Traps Distractor: “PaaS gives you full control over the underlying OS.” – Wrong; OS is provider‑managed. Distractor: “Public PaaS is always the cheapest option.” – Incorrect at large scale due to managed‑service premiums. Distractor: “Hybrid PaaS means you must buy two separate licenses.” – Misleading; it’s a single service with flexible deployment. Distractor: “All PaaS offerings support any programming language.” – False; language/framework support varies by vendor. --- Use this guide for a quick, confidence‑boosting review before your exam!