Since I’m preparing with my AZ-900 exam I will sharing some of my notes to explain some basic concept regarding cloud computing and demonstrate part of Azure service. This will be series of post with total 6 of chapter. And covered with latest skilled measured you can using it to preparing with AZ-900 exam if you want or as an introduction with Azure.
Covering the following topic:
- The benefits of cloud computing
- What are cloud service models?
- Cloud deployment model comparison
- The differences between CapEx and OpEx
- What is consumption model?
- What is shared responsibility model?
- What is serverless computing
What is Azure? Speaking of the “Cloud Computing” is not a fresh term as nearly all website that you are access right now via internet are hosted in the “cloud” somewhere else in the planet. The cloud is referred to a datacenter with hundreds thousands of server running with dedicated power supply and cooling device in a secured building. The company who providing these services is referred to as cloud provider.
When you starting to interact with cloud you first thing is to thinks about: Where you want your data to stored? Where you want your application to be run? How your employees and customers to access it? How many infrastructure you wish to manage?
There are three different cloud deployment models you should knowing that advantages and disadvantages before you start your deployment.
Public Cloud
Like the name implies your data stored on your service provider side and you do not have any of hardware to be managed, everything is runs on your cloud service provider’s hardware.
Advantages:
- You pays only for what you used, there no upfront costs.
- You don’t have to buy new server in order to scale or upgrade it.
- You are not responsible for maintenance or update of the hardware.
Disadvantages:
- You had specially requirements for your application or because it was legacy application that cloud service provide does not support it.
- You have some specific security concerns or requirements that you cannot meet.
- You may have some customer in government agencies that it have policies or legal requirements.
- Due to you don’t have hardware to manage that you are afraid of it may not go as you want to.
Private Cloud
A private cloud is usually you create a cloud environment in your own datacenter or hosted by a third-party service provider. This is feels like the same as public for your end users, but you remain completely responsible for the maintenance of the hardware and software. If that’s locate on-premise then you may have to manage it whole infrastructure.
Advantages:
- As it was hosted on your on-premise datacenter you have total control of your hardware and your data.
- It can meet your security and compliance requirements for your customers.
- You can ensure all hardware & software configured as you wish, also support your legacy application.
Disadvantages:
- You have upfront cost due to you must purchase the hardware for setup and maintenance.
- You lost agility to scale as you have to buy new hardware and perform the setup.
- You must have dedicated IT staff to be manage and monitor it.
Hybrid Cloud
A hybrid cloud is combines both public and private clouds, allowing you to run your applications in the most appropriate location, or more flexible. For example, your company most service and application are in the public cloud but there some highly confidential data that need to be kept in local then you can deploy with hybrid cloud let your local file server or database to work along with your application or service that hosted in the public cloud.
Advantages:
- You have completely control of your confidential data.
- You can keep your legacy application or hardware to be running along with rest of service that hosted in public cloud.
- You have the flexibility to chose which part of your service or application. that go public or stay at on-premise.
Disadvantages:
- It could be more complicate to setup and manage.
- It involve some upfront cost.
- May require dedicated IT staff to manage.
Now you know these three of cloud deployment models so what’s next? Say your company wanna to take local on-premise server and moving to the cloud, based on previously introduction you choose public cloud because it doesn’t have highly confidential data remain also you only have limited budget, so your next move is to choose the service model that fit your requirements.
There three major service models categories in cloud computing: IaaS / PaaS / SaaS. Let go check each of them starting with IaaS.
Infrastructure as a service (a.k.a IaaS)
Simple put Infrastructure as a service is you rent a hardware (typically a dedicated server) from your cloud service provider. Same as you purchased a new server and installed locally which gives you the most control of the hardware, but the difference is you have it instantly and disposal it when you are not need it. There is not upfront cost you pay as you go.
But one important thing to be remember here, when you using Iaas doesn’t mean you just leaves everything except the data to the cloud service provider. You have to maintain the hardware and the OS same as you did to on-premise server, and the cloud service provider will make sure that infrastructure you’re using are function correctly. That’s called shared responsibility model.
IaaS service model typically used in:
- Test and development
- Remotely work at a secured environment (e.g. Windows Virtual Desktop on Azure)
- Storage, backup, recovery
- Migrating workloads
Platform as a service (a.k.a PaaS)
Platform as a service basically providing you with a managed hosting environment that you can deploy your service or application quickly without managing the underlying infrastructure. Like you wish to host your company website you can choose PaaS hosted with Windows Server or Linux distribution without worrying about the OS configuration.
PaaS service model typically used in:
- Web hosting services
- Content Delivery Service
- Development framework
- Business intelligence
Software as a service (a.k.a SaaS)
Software as a service is the model where you usually using a software with a monthly or annually subscription through a web browser that you don’t have to download or install with a specific package. Like you using Office 365 or Google Doc with a web browser to write a document. In this model the end user does have to know the rest of underlying configuration except only to use it.
SaaS service model typically used in:
- Providing service directly via web browser
- Focus on the main feature or code
- Don’t need to manage infrastructure
Catch up with earlier case then you should pick the IaaS service model to migrate your on-premise workloads to the cloud, but don’t forget the shared responsibility when you choose IaaS.
What are the benefits of cloud computing?
- It’s cost-effective
You pay as you go and there is not upfront cost here. By the end of the day you no long using those resource you are not billed - It’s elastic
If your hosted website reach large amount of requests it can automatically allocate more compute power to handle and when these requests are reduced it de-allocate the additional compute resource to minimize cost. - It’s scalable
You can easily to increase or decrease the resource you want with cloud computing at any time. Cloud computing support both vertical and horizontal scaling.
Vertical scaling means adding resource to increase the power of an existing server. Like adding more CPU core or more GB of memory.
Horizontal scaling means adding more servers that function together as one unit. Like you add two or more server to handle http requests. - It’s reliable and support disaster recovery
The cloud service provider usually offer with data backup and disaster recovery to make sure your data is always safe. Like Azure providing with Availability Zones to dealing with nature disaster such scenarios you can maintain your mission-critical application at high-availability. - It’s agility
You can quickly deploy your application or service when there was a changes made. - It’s secure
Large cloud service provider like Azure will a highly secured building with 24/7 monitoring and dedicated security personnel to maintain a high level of physical security.
Capital Expenditure versus Operational expenditure
Before the cloud computing really get hot in the market the IT department usually will purchase new server or a lot for their new project at the beginning, that was good the prerequisite is you know the exactly how much you will cost on the project and you wish have a precise predict. But if your project is nearly closed and it’s demand for more server to be complete then you have to purchase again and in the day that total compute power are only utilized only 50% – 60% that you are waste your money, this is when you go with CapEx(capital expenditure). OpEx on the other hand is particularly appealing if the demand fluctuates or is unknown. So with OpEx the IT department does have to buy with a lot of server ahead, instead you need how many then you add to it and if you don’t need to can quickly reduce it to adapt your project needs.
What is Azure?
Azure is Microsoft’s cloud computing platform that have several services you can use to build, manage, and deploy your application around the globe with mins.
There so many services and features that Azure can provide:
- Compute
- Networking
- Storage
- Mobile
- Databases
- Web
- Internet of Things
- Big Data
- Artificial Intelligence
- DevOps
Personally I recommend you register with an Azure free account from https://azure.microsoft.com to park around all of features you want. Sign up an Azure free account you can get a 12months of popular free cloud service with AI, compute, storage, network etc., and $200 credit to spent for 30days. It will require a credit card to be authenticate no cost will be generate and you don’t have to worry about when you try with these services and exceed and receive a large bill, there is a free trial subscription will assigned along side with your Azure free account. Last there a spending limit to ensure you aren’t charged once you have used up your credits. And every time when you login Azure Portal you will be notified how many credits left.
Cloud computing service & storage service
There are two most common service that all cloud service provider offered: compute power & storage. Let’s check out two of them.
With compute service the major one is VM(virtual machine) but there are other two popular options: containers and serverless computing.
What are containers?
Container provide a consistent, isolated execution environment for applications. They’re similar to VMs but don’t require a guest operating system. Simply put, the VM is virtualize the hardware but container is virtualize the OS. The application and all its dependencies is packaged into a “container” and then a standard runtime environment is used to execute the app. So make is quickly to start.
What is serverless computing?
The serverless computing lets you run application code without creating, configuring, or maintaining a server. The core idea is that your application is broken into separate functions that run when triggered by some action. For example, you can build a serverless process when someone leaves a comments on your website you get an email notification.
The serverless model major difference from VMs and containers is that you only pay for the processing time used by each application as it executes. On the other hand, VMs and containers are charged while they’re running, even most of time the application on them are idle.
Storage service
This is the second common service that cloud provider offered. You can store any type of data like normally you put on a file server. The advantage to using cloud-based data storage is you can scale to meet your needs. Like Azure providing data storage in virtual disk with capability to storing up to 32 TB of data enough for you to handle large amount of data such as videos project.
There are so many storage service Azure offered I’ll take about this in my next few chapter. BTW, when you stuck go check with online documentation or looking for community support from Azure knowledge center / Microsoft Tech Community / Stack Overflow.
