edg3/experiment_software_csharp_12_net_8 · Datasets at Hugging Face

title stringlengths 3 46	content stringlengths 0 1.6k
20:401	After that, the following form will appear:
20:402
20:403	Figure 20.4: Creating a Kubernetes cluster
20:404	It is worth mentioning that you can get help simply by hovering over any with the mouse.
20:405	As usual, you are required to specify a subscription, resource group, and region. Then, you can choose a unique name (Kubernetes cluster name) and the version of Kubernetes you would like to use. For computational power, you are asked to select a machine template for each node (Node size) and the number of nodes. While for an actual application, it is recommended to select at least three nodes, let’s select just two nodes for our exercise in order to save our free Azure credit.
20:406	Moreover, the default virtual machine should also be set to a cheap one, so click Change size and select DS2 v2. Finally, set Scale method to Manual to prevent the number of nodes from being automatically changed, which might quickly burn through your free Azure credit.
20:407	The Availability zones setting allows you to spread your nodes across several geographic zones for better fault tolerance. The default is three zones. Please change it to two zones since we have just two nodes.
20:408	After implementing the preceding changes, you should see the following settings:
20:409
20:410	Figure 20.5: Chosen settings
20:411	Now you can create your cluster by clicking the Review + create button. A review page should appear. Confirm and create the cluster.
20:412	If you click Next instead of Review + create, you can also define other node types, and then you can provide security information, namely, a service principal, and specify whether you wish to enable role-based access control. In Azure, service principals are accounts that are associated with services you may use to define resource access policies. You may also change the default network settings and other settings.
20:413	Deployment may take a little while (10-20 minutes). After that time, you will have your first Kubernetes cluster! At the end of the chapter, when the cluster is no longer required, please don’t forget to delete it in order to avoid wasting your free Azure credit.
20:414	In the next subsection, you will learn how to install and use minikube, a single-node Kubernetes simulator, on your local machine.
20:415	Using minikube
20:416	The easiest way to install minikube is the usage of the Windows installer you can find in the official installation page: https://minikube.sigs.k8s.io/docs/start/.
20:417	During the installation you will be prompted on the kind of virtualization tool to use. If you already installed Docker Desktop and WSL, please specify Docker.
20:418	If you have a different operating system, please follow the default choices, instead.
20:419	The installation of Docker Desktop is explained in the technical requirements of Chapter 11, Applying a Microservice Architecture to Your Enterprise Application. Please note that both WSL and Docker Desktop must be installed and Docker must be configured to use Linux containers by default.
20:420	Once you have minikube installed, you must add its binary to your computer PATH. The easiest way to do it is to open a PowerShell console and run the following command:
20:421	$oldPath=[Environment]::GetEnvironmentVariable('Path', [EnvironmentVariableTarget]::Machine)
20:422	if ($oldPath.Split(';') -inotcontains 'C:minikube'){ '
20:423	[Environment]::SetEnvironmentVariable('Path', $('{0};C:minikube' -f $oldPath), [EnvironmentVariableTarget]::Machine) '
20:424	}
20:425
20:426	Once installed, your cluster can be run with:
20:427	minikube start
20:428
20:429	When you have finished working with the cluster, it can be stopped with:
20:430	minikube stop
20:431
20:432	In the next subsection, you will learn how to interact with your minikube instance or Azure cluster through Kubernetes’ official client, kubectl.
20:433	Using kubectl
20:434	Once you have created your Azure Kubernetes cluster, you can interact with it via the Azure Cloud Shell. Click on the console icon in the top right of your Azure portal page. The following screenshot shows the Azure Shell icon:
20:435
20:436	Figure 20.6: Azure Shell icon
20:437	When prompted, select the Bash Shell. Then you will be prompted to create a storage account, so confirm and create it.
20:438	We will use this shell to interact with our cluster. At the top of the shell there is a file icon that we will use to upload our .yaml files:
20:439
20:440	Figure 20.7: How to upload files in Azure Cloud Shell
20:441	It is also possible to download a client called Azure CLI and to install it on your local machine (see https://docs.microsoft.com/en-US/cli/azure/install-azure-cli), but, in this case, you also need to install all the tools needed to interact with the Kubernetes cluster (kubectl and Helm) that are pre-installed in Azure Cloud Shell.
20:442	Once you’ve created a Kubernetes cluster, you can interact with it through the kubectl command-line tool. kubectl is integrated into Azure Cloud Shell, so you just need to activate your cluster credentials to use it. You can do this with the following Azure Cloud Shell command:
20:443	az aks get-credentials --resource-group <resource group> --name <cluster name>
20:444
20:445	The preceding command stores the credentials that were automatically created to enable your interaction with the cluster in a /.kube/config configuration file. From now on, you can issue your kubectl commands with no further authentication.
20:446	If, instead, you need to interact with your local minikube cluster, you need a local installation of kubectl, but minikube installs it automatically for you.
20:447	In order to use the automatically-installed kubectl, all kubectl commands must be preceded by the minikube command and kubectl must be followed by --. Thus, for instance, if you wanted to run the following command:
20:448	kubectl get all
20:449
20:450	Then you would have to write the following:
20:451	minkube kubectl -- get all
20:452
20:453	In the remainder of the chapter, we will write commands that work on actual Kubernetes clusters such as Azure Kubernetes. Therefore, when using minikube, remember to replace kubectl with minikube kubectl -- in your commands.
20:454	If you issue the kubectl get nodes command, you get a list of all your Kubernetes nodes. In general, kubectl get <object type> lists all objects of a given type. You can use it with nodes, pods, statefulset, and so on. kubectl get all shows a list of all the objects created in your cluster. If you also add the name of a specific object, you will get information on just that specific object, as shown here:
20:455	kubectl get <object type><object name>
20:456
20:457	If you add the --watch option, the object list will be continuously updated, so you can see the state of all the selected objects changing over time. You can leave this watch state by hitting Ctrl + C.
20:458	The following command shows a detailed report on a specific object:
20:459	kubectl describe <object name>
20:460
20:461	All objects described in a .yaml file, say myClusterConfiguration.yaml, can be created with the following command:
20:462	kubectl create -f myClusterConfiguration.yaml
20:463
20:464	Then, if you modify the .yaml file, you can reflect all the modifications in your cluster with the apply command, as shown here:
20:465	kubectl apply -f myClusterConfiguration.yaml
20:466
20:467	apply does the same job as create but, if the resource already exists, apply overrides it, while create exits with an error message.
20:468	You can destroy all objects that were created with a .yaml file by passing the same file to the delete command, as shown here:
20:469	kubectl delete -f myClusterConfiguration.yaml
20:470
20:471	The delete command can also be passed an object type and a list of names of objects of that type to destroy, as shown in the following example:
20:472	kubectl delete deployment deployment1 deployment2...
20:473
20:474	The preceding kubectl commands should suffice for most of your practical needs. For more details, the Further reading section contains a link to the official documentation.
20:475	In the next subsection, we will use kubectl create to install a simple demo application.
20:476	Deploying the demo Guestbook application
20:477	The Guestbook application is a demo application used in the examples in the official Kubernetes documentation. We will use it as an example of a Kubernetes application since its Docker images are available in the public Docker repository, so we don’t need to write software.
20:478	The Guestbook application stores the opinions of customers who visit a hotel or a restaurant.
20:479	It is composed of a UI, and an in-memory database, based on Redis. Moreover, updates are sent to the master copy of the Redis database, which is automatically replicated in N read-only Redis replicas.
20:480
20:481	Figure 20.8: Architecture of the Guestbook application
20:482	The UI application can be deployed in Kubernetes as a Deployment, since it is memoryless.
20:483	The Redis master store is deployed as a single pod Deployment. We can’t implement it with an N-pods Deployment since we need sharding for parallelizing updates. However, we might have used a StatefulSet assigning a different data shard to each different master Pod. However, since this is your first Kubernetes exercise and since write operations should not be predominant, a single master database should suffice in the practical case of a single restaurant/hotel.
20:484	Since all slave copies contain the same data and consequently are undistinguishable, they can be implemented with a Deployment, too.
20:485	The whole application is composed of three .yaml files that you can find in the GitHub repository associated with this book (https://github.com/PacktPublishing/Software-Architecture-with-C-Sharp-12-and-.NET-8-4E).
20:486	Here is the code for the master storage based on Redis that is contained in the redis-master.yaml file:
20:487	apiVersion: apps/v1
20:488	kind: Deployment
20:489	metadata:
20:490	name: redis-master
20:491	labels:
20:492	app: redis
20:493	spec:
20:494	selector:
20:495	matchLabels:
20:496	app: redis
20:497	role: master
20:498	tier: backend
20:499	replicas: 1
20:500	template:

20:401

After that, the following form will appear:

20:402

20:403

Figure 20.4: Creating a Kubernetes cluster

20:404

It is worth mentioning that you can get help simply by hovering over any with the mouse.

20:405

As usual, you are required to specify a subscription, resource group, and region. Then, you can choose a unique name (Kubernetes cluster name) and the version of Kubernetes you would like to use. For computational power, you are asked to select a machine template for each node (Node size) and the number of nodes. While for an actual application, it is recommended to select at least three nodes, let’s select just two nodes for our exercise in order to save our free Azure credit.

20:406

Moreover, the default virtual machine should also be set to a cheap one, so click Change size and select DS2 v2. Finally, set Scale method to Manual to prevent the number of nodes from being automatically changed, which might quickly burn through your free Azure credit.

20:407

The Availability zones setting allows you to spread your nodes across several geographic zones for better fault tolerance. The default is three zones. Please change it to two zones since we have just two nodes.

20:408

After implementing the preceding changes, you should see the following settings:

20:409

20:410

Figure 20.5: Chosen settings

20:411

Now you can create your cluster by clicking the Review + create button. A review page should appear. Confirm and create the cluster.

20:412

If you click Next instead of Review + create, you can also define other node types, and then you can provide security information, namely, a service principal, and specify whether you wish to enable role-based access control. In Azure, service principals are accounts that are associated with services you may use to define resource access policies. You may also change the default network settings and other settings.

20:413

Deployment may take a little while (10-20 minutes). After that time, you will have your first Kubernetes cluster! At the end of the chapter, when the cluster is no longer required, please don’t forget to delete it in order to avoid wasting your free Azure credit.

20:414

In the next subsection, you will learn how to install and use minikube, a single-node Kubernetes simulator, on your local machine.

20:415

Using minikube

20:416

The easiest way to install minikube is the usage of the Windows installer you can find in the official installation page: https://minikube.sigs.k8s.io/docs/start/.

20:417

During the installation you will be prompted on the kind of virtualization tool to use. If you already installed Docker Desktop and WSL, please specify Docker.

20:418

If you have a different operating system, please follow the default choices, instead.

20:419

The installation of Docker Desktop is explained in the technical requirements of Chapter 11, Applying a Microservice Architecture to Your Enterprise Application. Please note that both WSL and Docker Desktop must be installed and Docker must be configured to use Linux containers by default.

20:420

Once you have minikube installed, you must add its binary to your computer PATH. The easiest way to do it is to open a PowerShell console and run the following command:

20:421

$oldPath=[Environment]::GetEnvironmentVariable('Path', [EnvironmentVariableTarget]::Machine)

20:422

if ($oldPath.Split(';') -inotcontains 'C:minikube'){ '

20:423

[Environment]::SetEnvironmentVariable('Path', $('{0};C:minikube' -f $oldPath), [EnvironmentVariableTarget]::Machine) '

20:424

}

20:425

20:426

Once installed, your cluster can be run with:

20:427

minikube start

20:428

20:429

When you have finished working with the cluster, it can be stopped with:

20:430

minikube stop

20:431

20:432

In the next subsection, you will learn how to interact with your minikube instance or Azure cluster through Kubernetes’ official client, kubectl.

20:433

Using kubectl

20:434

Once you have created your Azure Kubernetes cluster, you can interact with it via the Azure Cloud Shell. Click on the console icon in the top right of your Azure portal page. The following screenshot shows the Azure Shell icon:

20:435

20:436

Figure 20.6: Azure Shell icon

20:437

When prompted, select the Bash Shell. Then you will be prompted to create a storage account, so confirm and create it.

20:438

We will use this shell to interact with our cluster. At the top of the shell there is a file icon that we will use to upload our .yaml files:

20:439

20:440

Figure 20.7: How to upload files in Azure Cloud Shell

20:441

It is also possible to download a client called Azure CLI and to install it on your local machine (see https://docs.microsoft.com/en-US/cli/azure/install-azure-cli), but, in this case, you also need to install all the tools needed to interact with the Kubernetes cluster (kubectl and Helm) that are pre-installed in Azure Cloud Shell.

20:442

Once you’ve created a Kubernetes cluster, you can interact with it through the kubectl command-line tool. kubectl is integrated into Azure Cloud Shell, so you just need to activate your cluster credentials to use it. You can do this with the following Azure Cloud Shell command:

20:443

az aks get-credentials --resource-group <resource group> --name <cluster name>

20:444

20:445

The preceding command stores the credentials that were automatically created to enable your interaction with the cluster in a /.kube/config configuration file. From now on, you can issue your kubectl commands with no further authentication.

20:446

If, instead, you need to interact with your local minikube cluster, you need a local installation of kubectl, but minikube installs it automatically for you.

20:447

In order to use the automatically-installed kubectl, all kubectl commands must be preceded by the minikube command and kubectl must be followed by --. Thus, for instance, if you wanted to run the following command:

20:448

kubectl get all

20:449

20:450

Then you would have to write the following:

20:451

minkube kubectl -- get all

20:452

20:453

In the remainder of the chapter, we will write commands that work on actual Kubernetes clusters such as Azure Kubernetes. Therefore, when using minikube, remember to replace kubectl with minikube kubectl -- in your commands.

20:454

If you issue the kubectl get nodes command, you get a list of all your Kubernetes nodes. In general, kubectl get <object type> lists all objects of a given type. You can use it with nodes, pods, statefulset, and so on. kubectl get all shows a list of all the objects created in your cluster. If you also add the name of a specific object, you will get information on just that specific object, as shown here:

20:455

kubectl get <object type><object name>

20:456

20:457

If you add the --watch option, the object list will be continuously updated, so you can see the state of all the selected objects changing over time. You can leave this watch state by hitting Ctrl + C.

20:458

The following command shows a detailed report on a specific object:

20:459

kubectl describe <object name>

20:460

20:461

All objects described in a .yaml file, say myClusterConfiguration.yaml, can be created with the following command:

20:462

kubectl create -f myClusterConfiguration.yaml

20:463

20:464

Then, if you modify the .yaml file, you can reflect all the modifications in your cluster with the apply command, as shown here:

20:465

kubectl apply -f myClusterConfiguration.yaml

20:466

20:467

apply does the same job as create but, if the resource already exists, apply overrides it, while create exits with an error message.

20:468

You can destroy all objects that were created with a .yaml file by passing the same file to the delete command, as shown here:

20:469

kubectl delete -f myClusterConfiguration.yaml

20:470

20:471

The delete command can also be passed an object type and a list of names of objects of that type to destroy, as shown in the following example:

20:472

kubectl delete deployment deployment1 deployment2...

20:473

20:474

The preceding kubectl commands should suffice for most of your practical needs. For more details, the Further reading section contains a link to the official documentation.

20:475

In the next subsection, we will use kubectl create to install a simple demo application.

20:476

Deploying the demo Guestbook application

20:477

The Guestbook application is a demo application used in the examples in the official Kubernetes documentation. We will use it as an example of a Kubernetes application since its Docker images are available in the public Docker repository, so we don’t need to write software.

20:478

The Guestbook application stores the opinions of customers who visit a hotel or a restaurant.

20:479

It is composed of a UI, and an in-memory database, based on Redis. Moreover, updates are sent to the master copy of the Redis database, which is automatically replicated in N read-only Redis replicas.

20:480

20:481

Figure 20.8: Architecture of the Guestbook application

20:482

The UI application can be deployed in Kubernetes as a Deployment, since it is memoryless.

20:483

The Redis master store is deployed as a single pod Deployment. We can’t implement it with an N-pods Deployment since we need sharding for parallelizing updates. However, we might have used a StatefulSet assigning a different data shard to each different master Pod. However, since this is your first Kubernetes exercise and since write operations should not be predominant, a single master database should suffice in the practical case of a single restaurant/hotel.

20:484

Since all slave copies contain the same data and consequently are undistinguishable, they can be implemented with a Deployment, too.

20:485

The whole application is composed of three .yaml files that you can find in the GitHub repository associated with this book (https://github.com/PacktPublishing/Software-Architecture-with-C-Sharp-12-and-.NET-8-4E).

20:486

Here is the code for the master storage based on Redis that is contained in the redis-master.yaml file:

20:487

apiVersion: apps/v1

20:488

kind: Deployment

20:489

metadata:

20:490

name: redis-master

20:491

labels:

20:492

app: redis

20:493

spec:

20:494

selector:

20:495

matchLabels:

20:496

app: redis

20:497

role: master

20:498

tier: backend

20:499

replicas: 1

20:500

template: