Implementing DevOps with Microsoft Azure: Automate your deployments and incorporate the DevOps culture

Cloud computing is a type of computing that provides multitenant or dedicated computing resources such as compute, storage, and network to cloud consumers on demand. It comes in different flavors that includes cloud deployment models and cloud service models.

Cloud deployment models describe the way cloud resources are deployed, such as behind the firewall and on-premise exclusively for a specific organization, and this is called the private cloud; or cloud resources that are available to all organizations and individuals, and this is called the public cloud; or cloud resources that are available to a specific set of organizations that share similar types of interests or similar types of requirements, and this is called the community cloud; or cloud resources that combine two or more deployment models, and this is known as the hybrid cloud.

Cloud service models describe the way cloud resources are made available to cloud consumers. It can be in the form of a pure infrastructure, where virtual machines are accessible and controlled by a cloud consumer or end user, and this is called Infrastructure as a Service (IaaS); or a platform where runtime environments are provided so installation and configuration of all software needed to run application are already available and managed by a cloud service provider, and this is called Platform as a Service (PaaS); or Software as a Service (SaaS), where a whole application is made available by a cloud service provider with the responsibility of infrastructure and platform remaining with the cloud service provider.

Cloud computing has a few characteristics which are significant such as multitenancy, pay-as-you-use (similar to electricity or gas connection), on-demand self-service, resource pooling for better utilization of compute, storage; network resources, rapid elasticity for scaling up and scaling down resources based on needs in an automated fashion, and measured service for billing.

Over the years, usage of different cloud deployment models has varied based on use cases. Initially, the public cloud was used for applications that were considered non-critical while private cloud was used for critical applications where security was a major concern. Hybrid cloud usage evolved over time with experience and confidence in the services provided by cloud service providers.

In a similar way, there was a usage pattern for cloud service models too. As we usually do in a normal traditional environment, it was easier to adopt IaaS as there is complete control. Over time, organizations realized the cost of managing resources in the cloud as the efforts are the same in managing resources, considering security configurations, and other configurations. Hence, IaaS was becoming difficult day by day with the evolution of PaaS. PaaS has evolved over the years and its approach is much wider, and its services include multiple programming languages such as .NET, Java, PHP, Python, and Ruby.

In simple English, PaaS provides infrastructure as well as a runtime environment to deploy an application. The difference is the end user doesn't have control on infrastructure while we can configure a runtime environment and debug an application remotely and troubleshoot issues to some extent. There are PaaS offerings where you can have dedicated infrastructure resources for application deployment, but even in that case, control of infrastructure is in the hands of cloud service providers. In short, we as cloud users need not to worry about downloading different web servers, database servers, and other required software and install them on a machine and maintain them. We don't even need to worry about updating them to newer versions. Cloud service providers are responsible for maintaining availability of the application and its resources. However, as users, we need to follow best practices to configure applications to have a highly available application considering different services provided by the cloud service provider.

Comparison – PaaS and IaaS

Let's visualize how the process workflow is executed in terms of a traditional model or in IaaS, and then we will compare it with PaaS. In a traditional environment, the infrastructure provisioning process takes place in a different manner than the acquisition of a virtual machine in a cloud subscription.

While in the case of PaaS, the flow has less complications than the traditional or IaaS process.

However, the approval process in the cloud environment too has costs associated with it, so organizations keep their own sets of e-mail approval processes to create a virtual machine or to provision any PaaS offering.

Over the years, PaaS has gained its momentum and many organizations have realized that there are less complexities and less management overhead involved in it. Hence, many customers are leaning towards the PaaS offering, as they don't wish to manage virtual machines and apply security policies in the cloud and maintain all of them. In IaaS, as users, we need to configure high availability and scalability, and it is more complicated compared with PaaS.

PaaS offerings manage load balancers and high availability with little configuration and hence save a lot of time, and the architecture is more clear. We need to remember that most of the control is available with cloud service providers and hence we have less things to manage. Cloud service providers have more control and they implement all the best practices and a standard pattern to fulfill the SLAs attached with PaaS offerings. Governance wise, PaaS is less flexible in terms of applying policies as it is a shared environment and is standardized as well. Managing PaaS is also less costly compared with IaaS. In the case of IaaS, we need to manage everything the way we manage on-premise.

In short, those who know better about infrastructures and platforms manage them with efficiency so we have less overhead.

An Application Platform as a Service (aPaaS) provides features to design, develop, deploy, and manage an application life cycle, enabling effective resource utilization, agility, and faster time to market. It is getting popular now-a-days as it has less overhead and less complexities of managing resources for the application life cycle management. Usage of thick clients is an old story now and PaaS providers have realized the trend. Developers want offerings all in once place where they can develop, compile, package, and deploy at them. PaaS is accessed through web browsers such as Firefox, Chrome, or Internet Explorer. Most PaaS providers have come up with offerings to support developers within browser IDEs to develop applications. In recent times, DevOps has also gained momentum and PaaS along with the DevOps culture is breaking a lot of barriers in the traditional culture of IT. In the next section, we will discuss DevOps and then we will see benefits of PaaS in DevOps.

DevOps is more about the culture of an organization, processes, and technology to develop communication and collaboration between Development and IT Operations teams so as to manage the application life cycle more effectively than the existing ways of doing it. We often tend to work based on patterns to find reusable solutions from similar kind of problems or challenges.

Over the years, achievements and failed experiments, best practices, automation scripts, configuration management tools, and methodologies became an integral part of the culture. This helps to define practices for ways of designing, developing, testing, setting up resources, managing environments, managing configuration, deploying an application, gathering feedback, performing code improvements, and doing innovations. It helps to train people to think that drive the DevOps culture at every stage.

The DevOps culture is considered to be an innovative package to integrate the Dev and Ops teams in an effective manner that includes components, such as continuous build Integration, continuous testing, cloud resource provisioning, CD, continuous deployment, continuous monitoring, continuous feedback, continuous improvement, and continuous innovation, to make application delivery faster as per the demands of the agile methodology.

Considering the way we have been doing things to manage applications over the years, it is a challenge to change the culture. Mr. David Gleicher created the formula for change, and later it was refined by Kathie Dannemiller. This formula is still relevant as it provides a model to assess the relative strengths affecting the success possibilities of organizational change initiatives.

As per the formula, three factors must be available for a meaningful organizational change to take place. I just tried to classify different things in these three factors:

If the product of these three factors is greater than R = Resistance, then change is possible. If any one factor is absent (zero) or low while D, V, and F are multiplied, then the multiplication will be zero or low and therefore it may not overcome the resistance. Even without numerical values, the outcome of this formula is very much in the favor of change considering the values.

To bring an organization-wide change, we must consider the possibility of dissatisfaction in people and try to bring the change in mindsets by sharing industry trends, leadership ideas, best practices, and competitor analysis to identify the necessity for change. Hence, to ensure a successful change, it is the need of the hour to use influence and think strategically to create a vision and identify the basic steps toward it. To change the culture of an organization effectively, we the people need to bring in an agile, standardized environment, uniform automation processes orchestration, and DevOps enablers. Essentially, it means the combination of people, processes, and tools to achieve efficiency.

In other words, the DevOps culture is not much different than the organization culture, which has shared values and behavioral aspects. It needs adjustments in mindsets and processes to align with the new technology and tools.

At the end of the day, the ball is in our court and we should remember that we can't make an omelette without breaking a few eggs. To be more precise, the early bird catches the worm.

In the next section, we will describe in detail about the role and benefits of PaaS and aPaaS in DevOps.

App Service is one of the most popular offering from Microsoft Azure. It is a PaaS. There are four kinds of applications that are created in App Service:

In this book, our focus is on Web Apps. Web Apps are a PaaS offering that have computing resources and runtime environments managed by Microsoft Azure while the user is only responsible for the application and configurations related to Web Apps and high availability. The following are some quick points about Azure Web Apps:

Let's see some basic difference between Azure Virtual Machines and Azure Web Apps:

To have a quick hands on, follow these steps:

This is how we can create a sample Azure web app.

The Microsoft Azure SQL Database is a RDBMS in the cloud. It supports the SQL Server engine and hence we can use existing SQL Server tools, libraries, and APIs to manage the SQL Database in the cloud. It comes in different pricing tiers such as Basic, Standard, and Premium service tiers that have different capacities for different workloads.

Microsoft Azure Storage is a cloud storage offering that is highly available, scalable, and durable. Microsoft Azure Storage services provide Blob storage, Table storage, Queue storage, and File storage:

Azure Active Directory (Azure AD) is a cloud-based, multitenanted, and highly available identity management service from Microsoft. It can manage users, groups, and multi-factor authentication; add an application organization that is developing for authentication; add an application from the gallery for authentication; add a custom domain; add role-based access control; and so on.

To access Azure AD from the Azure portal, go to https://portal.azure.com and click on Azure Active Directory.

Verify the Overview details with Users and groups, App registrations, and Azure AD Connect.

The application Gallery supports 2,771 applications for Azure AD integration at the time of writing this chapter. Categories include Business Management, Collaboration, Construction, Content management, CRM, Data Services, Developer Services, E-commerce, Education, ERP, Finance, Health, Human resources, IT infrastructure, Mail, Marketing, Media, Mobile Device Management, Productivity, Project Management, Security, Social, Supply Management, Telecommunications, Travel, and Web Design and Hosting.

Major applications supported are Office365, Salesforce, ServiceNow, Google Apps, and so on. Go to https://manage.windowsazure.com to access Azure AD from the classic portal:

One of the biggest benefits that we will utilize in this book is the way we can provide authentication to access Azure Web Apps. We can configure Azure Web Apps with Directory so only specific users can access the application hosted on Azure Web Apps.

Visual Studio Application Insights is a flexible analytics service. It helps us to get the insights of performance and usage of an application. It can be used for .NET- or J2EE-based applications that are hosted on-premise or in the cloud. We will cover only a few important features that come with this service in this book.

Create a sample application and go to its MONITORING section. Click on Application Insights. Select Create new and click on OK:

We can create a web test for testing the availability of applications from multiple regions. We can select a ping test or multi-step test to check the availability. The alert criteria can also be configured. Performance testing is also a very interesting feature available in App Insights. It is more of a load testing based on the number of users for a specific duration.

We will see the how to of both in this book where we intend to cover monitoring.

Microsoft Azure Traffic Manager provides a feature to distribute user traffic to different endpoints. These endpoints can be Azure App Service (Azure Web Apps), cloud services, Azure Virtual Machines, and external endpoints. It is a DNS-based traffic routing. Azure Traffic Manager supports three traffic routing methods to decide how traffic can be routed to different endpoints. In simple terms, it is a way to decide which endpoint should serve the DNS request.

There are two different deployment models in Microsoft Azure—the classic deployment and Resource Manager deployment models. Microsoft Azure Traffic Manager uses different terminology for traffic routing methods (known as the load-balancing method in the classic deployment model):

Often this method is cost-optimized and the usage of the application is very well known and specific to a region.

Let's say we know that our application hosted in Azure Web Apps is going to be equal or in a specific ratio, then the Weighted method can be more appropriate.

So, in simple terms, users will be redirected to the nearest endpoint to serve the response. Some benefits of Microsoft Azure Traffic Manager are as follows:

Microsoft Azure services are available in 34 regions around the globe and more regions are continuously planned to be supported. The more regions, the more it allows customers to achieve better performance with cost optimization. It also helps in scenarios where data location is legally restricted.

To verify WEB + MOBILE products available by region go to https://azure.microsoft.com/en-in/regions/services/ and check the WEB + MOBILE section:

Azure is generally available in 34 regions and 12 geos around the globe. It has already announced plans for six additional regions and two additional geos. For customers, it is extremely important to have legal compliance in the context of a storage location of their data. There are two different possibilities/authorities in this scenario:

Resource groups in Microsoft Azure is nothing but a logical container. It can be used to group all different resources such as App Service, SQL Databases, and Storage Accounts, available in Microsoft Azure. We are going to consider services that we will use in this chapter for most of the examples. Resource groups provide a simple way to manage resources together. One of the biggest advantages is to manage the role-based access on the resources in an easy manner.

For example, we need to create resources such as Azure Web Apps, SQL Database, and Storage Account in the West US and provide access to all of them to some users. It is painful to assign a user to individual resources. Rather, it is more manageable if we can provide group access to all resources. This way resources can be managed in a better way.

To create a resource group, click on Resource groups in the left-hand sidebar menu. Click on the +Add button to create a resource group. Provide the Resource group name, select Subscription, select Resource group location, and click on Create:

Wait until the resource group is created.

Once the eTutorialsWorld resource group is created, click on it and verify the Overview section.

As of now, there are no resources in the resource group; hence there are No deployments in the Overview section—No resources to display:

As there are no resources, there is no cost available in the Resource costs section of the eTutorialsWorld resource group:

We will use this resource group in the coming chapters as a logical container for different resources such as Azure Web Apps and Azure SQL Database.

An App Service Plan (ASP) is a set of capacities (the instance size and instance count on which the application is hosted) and features. Capacity is directly linked to cost and hence it is similar to choosing a pricing tier. There are different capabilities and limits within ASPs.

There are five pricing tiers, namely Free, Shared, Basic, Standard, and Premium. Each ASP can be used for different purposes and they are different in providing features too. SLAs for Basic, Standard, and Premium are 99.95 percent. Autoscale, geo-distributed deployment, VPN hybrid connectivity, deployment slots, and automated backups are available only in the Standard and Premium tiers. In the Standard tier, five deployment slots are available, while in the Premium tier, 20 slots are available. Another major difference is the maximum instances, as they are directly associated with scaling. Hence, it is important to pick the proper tier or ASP to gain the desired performance. The Basic, Standard, and Premium tiers allow up to 3, 10, and 50 instances respectively.

The following are some important points regarding ASPs:

Let's consider the Azure calculator and understand how the pricing works. Go to https://azure.microsoft.com/en-in/pricing/calculator/. Click on +Add items and click on App Service in the Featured category:

Select REGION, TIER, INSTANCE SIZE; by default, the instance size is 1 and the hours are 744. Verify the cost estimate:

Now change the instance count and verify the cost:

As the instance count directly affects pricing, let's see from where we can change the instance size as we need to be careful while doing it.

Let's create an ASP and verify the instance count. Go to the left-hand side menu bar, find App Service plans, and click on +Add.

Select the eTutorialsWorld resource group we created; select Operating System, Location, and Pricing tier. Click on Create:

Once the ASP is created, verify its Overview section. Verify the Apps / Slots count. It is 0 / 0 as no app or the slot is using the ASP:

Go to the Scale out (App Service plan) section in APP SERVICE PLAN, and verify the default number of instances. As we kept the Basic pricing tier, it allows us to scale up to three instances only:

If we want to cancel the changes, click on Discard.

Scaling resources is a significant part of making an application highly available and having good performance. There are two types of scaling:

Horizontal scaling is the most commonly used as we can schedule it also. It is important to understand what will impact on the performance and availability of an application and accordingly we can decide the scaling part:

There are three types of scaling out supported in Azure Web Apps:

Scale instances manually

We can configure a number of instances manually based on the existing patterns and usage and these will serve the application hosting. We cannot change this dynamically. We need to change the setting in the ASP:

 

Scale instances by CPU percentage

We can automatically or dynamically scale up or down instances based on the performance of the CPU we have configured. We can also configure the number of instances too.

 

Scale instances by schedule and performance rule

Based on multiple rules, we can increase or decrease the number of instances used in an ASP. This is the most flexible and widely used scale out option out of the three: