Azure’s compute offerings fall into three main categories: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS).
SaaS ( Software-as-a-Service)
Software as a service (SaaS) allows users to connect to and use cloud-based apps over the Internet. Common examples are email, calendaring, and office tools (such as Microsoft Office 365) so applications are built and hosted through 3rd party vendors who typically charge for a certain level of service.
SaaS provides a complete software solution that you purchase on a pay-as-you-go basis from a cloud service provider (example Office 365). You rent the use of an app for your organization, and your users connect to it over the Internet, usually with a web browser. All of the underlying infrastructure, middleware, app software, and app data are located in the service provider’s data center. The service provider manages the hardware and software, and with the appropriate service agreement, will ensure the availability and the security of the app and your data as well. SaaS allows your organization to get quickly up and running with an app at minimal upfront cost.
PaaS ( Platform-as-a-Service)
Platform as a service (PaaS) is a complete development and deployment environment in the cloud, with resources that enable you to deliver everything from simple cloud-based apps to sophisticated, cloud-enabled enterprise applications. You purchase the resources you need from a cloud service provider on a pay-as-you-go basis and access them over a secure Internet connection.
Like IaaS, PaaS includes infrastructure—servers, storage, and networking—but also middleware, development tools, business intelligence (BI) services, database management systems, and more. PaaS is designed to support the complete web application lifecycle: building, testing, deploying, managing, and updating.
PaaS allows you to avoid the expense and complexity of buying and managing software licenses, the underlying application infrastructure and middleware or the development tools and other resources. You manage the applications and services you develop, and the cloud service provider typically manages everything else.
In short, IaaS gives you a server in the cloud (virtual machine) that you have complete control over. With an Azure VM, you are responsible for managing everything from the Operating System on up to the application you are running. This mode of operation will feel most like a typical on premises virtual machine where you remote desktop into the server to manage it instead of sitting down in front of a physical keyboard and mouse. If you need a solution that requires custom third party software or multiple applications running on a single machine, then IaaS might be for you.
This is an image from the Azure documentation that shows the generally available and coming-soon regions in the world. One beauty of the region is that you can geodistribute your application. You can create fault tolerance and load balancing by leveraging Azure Traffic Manager. You can also setup hot or warm standby disaster recovery sites using regions. They’re very flexible.
If you scroll down you will find Products by region. This allows you to cut down on the regions that you’re looking and find out which products are available in your region.
Here we can quickly check any of the regions, America, Europe, or Asia Pacific, and we have an icon for good, warning, error, and information. You can also get a quick check of Azure’s service status from within the Azure portal by looking at the service health widget. If you right-click inside that widget, you’ll notice there’s an option to go to the Azure status page.
AZURE LATENCY TEST
When you working with Azure there are a number of pages that you need to have at your reference at all times. Azurespeed.com is one of those. You see we have Please select region, check or uncheck all. I like to uncheck all. And then depending on where you are in the world you might be wondering (like I used to wonder actually)what region is my best choice. You get a live running list of latency measures, and it gives you your closest datacenters in a millisecond time. And you see that changes. And you’ll want to look at this over time. That’s why the graph down is helpful because it gives you a running list where you can take a look over the last several seconds or minutes and look for trends.
Service Level Agreements
Something you have to know before deploying stuff to cloud is how SLA works in Azure.
Here you can read all about SLA for VMs, storage etc.
So why are SLAs important in the first place? It’s an agreement between an IT service provider and the customer. That’s you and I. The SLA describes the IT service, documents the service level targets, and specifies the responsibilities of the IT service provider and the customer. And this is an ITIL definition. When you move to the cloud, you’re outsourcing. That’s a simple fact of life. You can call it cloud or you can call it data center hosting, but we’re outsourcing portions of the technology stack, in this instance, to Microsoft. So Microsoft service level agreements, and indeed their service level descriptions, describe in quite a lot of detail the service that you’re actually going to achieve. SLA and SERVICE DESCRIPTIONS define what you can expect from Microsoft. You’re going to get it wrong. It’s critical that you don’t put anything into production without having a service level backed configuration or supported configurations. It’s very easy just to assume that the things that you know in the physical and the virtual world are the same in the Microsoft Azure world.
What do you need to know about the SLAs in Azure??
• Each service has its own unique SLA
• When combininb SLA’s you have to use the lowest common denominator.
• STORAGE –> Managed Disks, Storage Accounts, Recovery Vaults
• NETWORK –> VPN & VPN Gateway, NSG, Public IP
VIRTUAL MACHINE SLA
There are only 2 SLA’s
• 99.5 % for Single VM
• 99.95% for 2 or more vms in availability set
This single virtual machine 99.5% SLA is only if your virtual machine is configured on premium storage. Flip that on its head for a moment. If you have a single machine and it does not have premium storage, it has standard storage, then technically you’re not running on an SLA at all. So no premium storage equals no SLA. Bear in mind you do have options, and one of those options is to do nothing and rely on good, solid backup.
Digging deeper into that 99.95%, Microsoft will guarantee that there is at least one virtual machine available for 99.95% of the time. So if you have two web servers, at least one of them, if configured in an availability set, will be available 99.95%. Assuming my two web servers are providing the same service, that’s really good news. But if those virtual machines are in an availability set but actually provide different services, well if one is up and the other one is down, arguably Microsoft have delivered the SLA, but your application is down.
So it’s really important to only strive, I guess, for that higher SLA if the application truly can make use of high availability. So if you can build a high availability layer into your application, let’s say for example the SQL layer or at the application layer or at the web or presentation layer, then you can put them into an availability set, and you’re going to get that 99.95. Flip that around, if you have a single web server, a single application server, and a single SQL server, and you decide to put them all in an availability set, yes those three machines technically between them will deliver 99.95, meaning that at least 1 of them will be up, but your standard 3-tier application will effectively be offline. The SQL might be up and everything else is down.
When though can you really rely on 99.95? Domain controllers, Active Directory, it’s a multi-master environment. Domain controllers work well together, so you can put them in an availability set, and at least one will be available 99.95%. SQL AlwaysOn clusters. Again, to nodes in a SQL AlwaysOn cluster, great to know that at least one of them will be available. And I mentioned web servers, you know the common network load balanced.
The SLA from a managed disk is actually based on the underlying storage. Now ironically in managed disks, you don’t get to see the storage and you don’t get to configure it, but you get to choose the type. So if you’re choosing a managed premium disk, then that is open to that single virtual machine’s service level agreement. If you’re taking a managed standard disk, i.e. not premium, then the virtual machine that is using that standard managed disk, that single virtual machine, is not subject to a service level agreement.
Get Started with Azure
Browse to https://azure.microsoft.com and click on Start Free
The way the evaluation process works with Azure is that you’ll receive 200 dollars in service credit that you can apply to any Azure service. The limitation is that it expires in one month, so you’ll want to get started kicking the tires, so to speak, as quickly as you can. To complete the registration process, you’ll need a free Microsoft account ID. This is the same Microsoft account that you can use across many different Microsoft services, including OneDrive, Office 365, Xbox Live, and so forth. You’ll also need a credit card. Now, note that the credit card is used simply for identity verification. Over the course of the month, you’ll not be charged at all because you’ll be using your $200. If you go through the $200 within a month, then any services you have running will stop, but you will not be charged, okay? Even at the end of your trial, your account will not be automatically switched to a pay-as-you-go model. That should give you some comfort in terms of that fear of an unexpected charge at the end of the month. Once you’ve finished setting up your account, you’ll be taken to the Microsoft Azure portal.
HOW TO BUY AZURE SERVICES
We have 2 options:
- Pay-as-you-Go –> most common one and the benefit of Pay-as-You-Go is that there’s no minimum spend. You pay only for what you use, and you can cancel your subscription at anytime. Another fear that some businesses have when they’re considering a public cloud is what’s called vendor lock-in, and that’s not the case with Azure.
- Enterprise Agreement, or EA, with Microsoft –> The idea with EA is that you’re prepaying for Azure services. There’s an upfront investment on your part. CAPEX rather than OPEX. The advantage to that, you get a little discount, but the main advantage is being able to more easily plan for allocation, instead of reacting to a bill. You also are able to take that pool of allocation, and spread it across multiple Azure services, like IaaS VMs, OMS, and so forth. Finally, EA customers get access to a special Azure Enterprise Portal which makes it easier to manage multiple subscriptions and resources.
The Monthly Spend – Points to Ponder
Number one, if you are using a trial subscription to Azure, please take comfort that Microsoft does not automatically convert that trial subscription to a Pay-as-You-Go plan. You might remember that you are required to provide an honest to goodness credit card when you create your account, but that’s done for identity verification, not for spending. You have to explicitly convert your trial account to either Pay-as-You-Go or Enterprise Agreement for the resources to begin accruing real cost.
Another thing we can do is set billing alerts. If you’ve forecasted, and determined X amount per monthly spend, then you can set alerts to make sure that you and your colleagues are notified when your spending reaches a particular level. You can actually put a spending limit on your subscription, such that your services will stop if you exceed a particular monetary threshold. Understand however that that may not be a good thing if you’re in a production situation, and you’re subject to SLAs.
Now when we know all of that we are ready to move to the next part which is dealing with subscriptions.
Thanks for reading!