Promises in writing

High availability and SLAs

High availability means keeping a service available as much as possible, regardless of disruptions. Azure is a highly available cloud with uptime guarantees that depend on the service. Those guarantees are part of each service-level agreement (SLA). Every extra nine in the percentage cuts the allowed downtime a lot.

SLA       Max downtime per year
99.9%     about 8.8 hours
99.95%    about 4.4 hours
99.99%    about 53 minutes
99.999%   about 5 minutes

Exam tip You get high availability by design: more than one instance, in more than one place. A single VM is a single point of failure.

Scalability: up or out

Scalability is the ability to adjust resources to meet demand. Vertical scaling changes a resource's power: scale up by adding processors (CPUs) or memory (RAM) to a virtual machine (VM), or down by removing them. Horizontal scaling changes the number of resources: scale out by adding VMs or containers, or in by removing them, automatically or by hand. Because the cloud is consumption-based, scaling down or in also cuts your bill.

Rosa: So on Black Friday you scale out, and in January you scale back in?

Dev: And you pay for the extra servers only while they're running.

Exam tip Up/down = bigger or smaller (vertical). Out/in = more or fewer (horizontal). Exam questions swap these on purpose.

Reliability

Reliability is the ability of a system to recover from failures and continue to function. It's one of the pillars of the Microsoft Azure Well-Architected Framework. The cloud's decentralized design helps: you can deploy resources in regions all over the world, so even if one region has a catastrophic event, other regions keep running. You can design apps to use this, and in some cases Azure shifts to another region automatically.

Predictability: performance and cost

Predictability lets you plan with confidence, and comes in two kinds. Performance predictability means predicting the resources needed for a good customer experience; autoscaling, load balancing and high availability support it. Cost predictability means forecasting your cloud spend: track resource use in real time, spot trends, and estimate future costs with tools like the Azure pricing calculator.

Mr. Barkley: Predictable bills. Now you're speaking my language.

Exam tip "Load balancing and autoscaling keep response times steady" = performance predictability. "Forecast next quarter's bill" = cost predictability.

Security: choose your level of control

The cloud lets you pick a security model that fits. Want maximum control? Infrastructure as a service (IaaS) gives you the operating systems and installed software to manage, patches included. Want patches handled for you? Platform as a service (PaaS) or software as a service (SaaS) take care of them automatically. Cloud providers are also well suited to handle distributed denial of service (DDoS) attacks, making your network more robust.

Governance at scale

Governance means making sure what gets deployed meets your standards and rules. Templates help ensure new resources meet your technical standards and regulatory requirements, and when standards change you can update resources at scale. Cloud-based auditing flags resources that are out of compliance with your baseline and suggests fixes. Depending on the service, patches and updates may even apply automatically.

Exam tip The governance tools themselves (Azure Policy, resource locks, Microsoft Purview) come later. Here, know the benefit: standards enforced and audited at scale.

Management of the cloud

Manageability comes in two kinds. Management of the cloud is about managing your resources. In the cloud you can:

• Automatically scale resource deployment based on need
• Deploy resources from a preconfigured template instead of setting them up by hand
• Monitor resource health and automatically replace failing resources
• Receive automatic alerts based on configured metrics, in real time

Management in the cloud

Management in the cloud is about how you reach and run your environment. You can manage Azure through a web portal, a command-line interface (CLI), application programming interfaces (APIs) or PowerShell. A team might deploy from templates, watch health in the portal, and automate routine jobs with CLI or PowerShell scripts.

Priya: So an alert at 2 a.m. instead of an angry customer at 9 a.m.?

Dev: That's the idea.

Exam tip "Of" = what the cloud does for your resources (scale, templates, replace, alert). "In" = the tools you drive it with (portal, CLI, APIs, PowerShell).