How I prepared for the AZ-900 and why it’s the certification where you should start with Azure

我是如何准备 AZ-900 考试的，以及为什么它应该是你开启 Azure 认证之旅的起点

In 2023, when I prepared for the AZ-900, I was studying Computer Science and had limited hands-on experience with cloud infrastructure. What stood out when studying was the exam’s scope: it doesn’t start with Azure services but with cloud computing fundamentals, which makes it useful for both technical and non-technical profiles. 2023 年，当我准备 AZ-900 考试时，我正在攻读计算机科学专业，对云基础设施的实践经验有限。学习过程中最让我印象深刻的是考试的范围：它不是从 Azure 服务开始，而是从云计算基础知识入手，这使得它对技术和非技术背景的人员都非常有用。

The AZ-900 (Microsoft Azure Fundamentals) is the introductory certification to the Azure ecosystem. It doesn’t require prior tech experience, doesn’t assume knowledge of programming, and doesn’t expect you to have administered a server. It’s designed for anyone who wants to understand how Microsoft’s cloud works, both from the technical side and the organizational and financial one. AZ-900（Microsoft Azure 基础知识）是 Azure 生态系统的入门级认证。它不需要先前的技术经验，不预设编程知识，也不要求你有管理服务器的经验。它专为任何想要了解微软云如何运作的人而设计，涵盖了技术层面以及组织和财务层面。

How I got to this exam: I got access through a program run by Microsoft and the Fundación Tomás Alva Edison in Mendoza, Argentina. The foundation has a partnership with Microsoft to bring these certifications to students and young people in the region. Today the exam is available through two channels. Most people book it through Pearson Vue, either in person at an authorized testing center or remotely from home through its OnVUE system. Students and educators can also book it through Certiport, which was the platform I used myself back in 2023 and remains active today as an option for that audience. 我是如何参加这场考试的：我通过微软与阿根廷门多萨的托马斯·阿尔瓦·爱迪生基金会（Fundación Tomás Alva Edison）合作的一个项目获得了考试机会。该基金会与微软建立了合作伙伴关系，旨在为该地区的学生和年轻人提供这些认证。如今，考试可以通过两个渠道进行。大多数人通过 Pearson Vue 预约，既可以在授权考试中心进行线下考试，也可以通过其 OnVUE 系统在家远程参加。学生和教育工作者也可以通过 Certiport 预约，这也是我在 2023 年使用的平台，目前它仍然是该群体的一个有效选择。

I’m a Microsoft Student Ambassador, and part of what I do in that role is help other students work through this same path. That’s why I put together a study plan on Microsoft Learn that organizes all the official modules in the order I’d recommend going through them: AZ-900 Study Plan. You can start it right now while you read this. 我是一名微软学生大使（Microsoft Student Ambassador），我的职责之一就是帮助其他学生走上同样的道路。因此，我在 Microsoft Learn 上整理了一份学习计划，按照我建议的顺序组织了所有官方模块：AZ-900 学习计划。你现在就可以在阅读本文的同时开始学习。

What the exam covers: The AZ-900 has between 40 and 60 questions, requires a minimum score of 700 out of 1000 to pass, and you’ll have 45 minutes to complete the assessment. There are no hands-on labs or case studies, and there’s no penalty for wrong answers, so it’s always worth answering even when you’re not sure. The cost is $99 USD in the United States, though Microsoft adjusts pricing by country. In Argentina, the official price is $59 USD. Check the exact price for your region on the official scheduling page before booking. 考试内容：AZ-900 包含 40 到 60 道题目，满分 1000 分，最低需 700 分才能通过，考试时间为 45 分钟。考试中没有动手实验或案例研究，答错不扣分，所以即使不确定答案，也一定要作答。在美国，考试费用为 99 美元，但微软会根据不同国家调整价格。在阿根廷，官方价格为 59 美元。在预约前，请务必在官方预约页面查看你所在地区的具体价格。

The content is organized into three major areas. Cloud concepts: The first area covers foundational concepts and is the most theoretical of the three. Before getting into Azure services specifically, the exam evaluates whether you understand what the cloud is and why it exists. 考试内容分为三大领域。云概念：第一个领域涵盖基础概念，是三者中最具理论性的部分。在深入了解具体的 Azure 服务之前，考试会评估你是否理解什么是云以及它为何存在。

The consumption model: One of the most important concepts is the difference between CapEx and OpEx. CapEx (Capital Expenditure) is the upfront investment in physical infrastructure you own: you buy servers, set them up, maintain them, and that cost stays with you for the life of the equipment. OpEx (Operational Expenditure) is the operating expense model: you pay for what you use, when you use it, with no upfront hardware investment. The cloud operates under OpEx, and that shifts the financial equation entirely, especially for smaller organizations. 消费模型：最重要的概念之一是 CapEx（资本支出）和 OpEx（运营支出）之间的区别。CapEx 是对你拥有的物理基础设施的前期投资：你购买服务器、进行设置和维护，这些成本在设备的使用寿命内一直存在。OpEx 是运营费用模型：你为所使用的资源付费，按需付费，无需前期硬件投资。云服务基于 OpEx 模式运行，这彻底改变了财务等式，对于小型组织尤其如此。

The service models: IaaS (Infrastructure as a Service) gives you the highest level of control. You rent the infrastructure (virtual servers, networking, storage) and you manage everything running on top: the operating system, middleware, applications, and data. A virtual machine on Azure is the clearest example. Think of it like renting an empty apartment where the building is already there but you decide everything that goes inside. 服务模型：IaaS（基础设施即服务）为你提供最高级别的控制权。你租用基础设施（虚拟服务器、网络、存储），并管理其上运行的一切：操作系统、中间件、应用程序和数据。Azure 上的虚拟机就是最明显的例子。可以把它想象成租了一间空公寓，大楼已经建好，但你决定里面放什么东西。

PaaS (Platform as a Service) removes the responsibility for the operating system and underlying infrastructure. You get a ready-to-use environment for building and deploying applications. Azure App Service fits here: you push your code and the platform handles the rest. It’s like working at a coworking space where the desk, internet, and infrastructure are already set up and you just show up to work. PaaS（平台即服务）免除了你对操作系统和底层基础设施的责任。你获得了一个用于构建和部署应用程序的现成环境。Azure App Service 就属于此类：你推送代码，平台处理其余事务。这就像在联合办公空间工作，办公桌、网络和基础设施都已经准备好，你只需要去工作即可。

SaaS (Software as a Service) is the model most people use every day without thinking about it. Microsoft 365, Teams, Gmail, and Notion are all SaaS. You use a finished application without thinking about any of the infrastructure underneath. SaaS（软件即服务）是大多数人每天都在使用却不自知的模型。Microsoft 365、Teams、Gmail 和 Notion 都是 SaaS。你使用成品应用程序，而无需考虑底层的任何基础设施。

Deployment models: Public cloud is infrastructure the provider shares across multiple customers. Private cloud is infrastructure dedicated exclusively to a single organization. Hybrid cloud combines both, which is especially useful for organizations with legacy systems they can’t fully migrate. The study plan covers all of this with hands-on exercises where you can practice without needing your own subscription. 部署模型：公有云是提供商在多个客户之间共享的基础设施。私有云是专门供单个组织使用的基础设施。混合云结合了两者，对于拥有无法完全迁移的遗留系统的组织特别有用。学习计划涵盖了所有这些内容，并配有动手练习，你无需拥有自己的订阅即可进行实践。

Azure architecture and services: This is the most extensive area of the exam and the one that covers the greatest number of individual services. It’s also the one that requires the most study time. Azure 架构和服务：这是考试中最广泛的领域，涵盖了最多的独立服务。这也是需要花费最多学习时间的领域。

Regions and availability zones: Azure is physically distributed across more than 70 regions worldwide. A region is a geographic area containing one or more datacenters. When you create a resource, you choose which region it lives in, and that choice affects latency, cost, and your organization’s data residency requirements. Within many regions there are availability zones, which are physically separate datacenters within that same region, connected by low-latency networks. If one zone fails, the others keep running. Azure also has region pairs: each primary region has a secondary region that’s geographically distant, for disaster recovery at a larger scale. 区域和可用性区域：Azure 在全球物理分布在 70 多个区域。区域是一个包含一个或多个数据中心的地理区域。当你创建资源时，你需要选择它所在的区域，这一选择会影响延迟、成本以及组织的合规性数据驻留要求。在许多区域内都有可用性区域，它们是同一区域内物理上独立的数据中心，通过低延迟网络连接。如果一个区域发生故障，其他区域仍能继续运行。Azure 还有区域对：每个主要区域都有一个地理位置遥远的次要区域，用于更大规模的灾难恢复。

Subscriptions and resource groups: Every Azure resource lives inside a well-defined hierarchy. The top level is the Microsoft Entra ID tenant, which represents your organization. Within the tenant there are subscriptions, which are the billing unit. Within each subscription you create resource groups, which are logical containers that group related services for a project or application. Every Azure resource always belongs to exactly one resource group. 订阅和资源组：每个 Azure 资源都存在于一个定义明确的层次结构中。顶层是 Microsoft Entra ID 租户，代表你的组织。租户内是订阅，即计费单位。在每个订阅内，你创建资源组，它们是逻辑容器，用于将项目或应用程序的相关服务分组。每个 Azure 资源始终只属于一个资源组。

Compute: Azure Virtual Machines are the classic IaaS model: a full server in the cloud that you manage from the operating system upward. They’re the most flexible. 计算：Azure 虚拟机是经典的 IaaS 模型：云中的完整服务器，你从操作系统层面开始进行管理。它们是最灵活的。