Exploration of AWS and azure in cloud computing (4)

Exploration of AWS and azure in cloud computing (4)

-- Amazon EC2 and Windows azure Virtual Machine

 

Next let's take a look at the creation of azure VM. The creation of virtual machines in azure is much simpler than that in AWS, with fewer configurations and shorter creation processes.

Create a virtual machine

First, go to the azure Management Portal.

 

Click the create button below and choose computing> Virtual Machine> Create from library

 

The creation method is the same as that of Amazon. It is also a wizard. The wizard interface is more modern. The first step is to select the operating system, including Windows Server 2008/2012, openlogic centos, SuSE Linux and Ubuntu, and BizTalk Server 2013 and SQL Server 2012.

 

Here we select Windows Server 2012. The next step is to enter the machine name of the virtual machine. The user name is administrator by default in Windows, and enter the Machine Password. Finally, select the instance type, that is, super small, small, medium, large, and super large. We select medium as AWS does.

 

Next, select the machine type, whether it is a separate virtual machine, or connect to an existing virtual machine. In Windows Azure, there is a cloud service concept. A cloud service has a public DNS name. You can create a new cloud service for each virtual machine, you can also place multiple virtual machines in the same cloud service and access them through Server Load balancer. The DNS name is the public name of this cloud service and can be used for global access. It must also be unique. The storage account can create or use existing files, which are used to store the file of the system disk image disk of the virtual machine, that is, the location of the vhd file. We will further describe the concept of a storage account when comparing S3 and azure storage accounts in the future. Next is the region/geographic group/and virtual network. There are several concepts here. The first region is the region of the azure data center we have discussed earlier. There are currently six regions, and there will soon be the seventh in mainland China. The affinity group in the geographic group and available zone in AWS want to be similar, and it also ensures that the machine is in the same machine room. What is different is that AWS pre-defines all AZS, and azure Ag names can be defined by themselves, which can ensure that they are in the same data center but not in any location. Finally, virtual networks are a concept of network configuration in azure. Different virtual machines can be placed in different subnet definitions, and hybrid clouds can be constructed through VPN. Therefore, you can also define the virtual network in which the virtual machine is put. If you select a virtual network, you also need to select the corresponding subnet. You do not need to select a region or a geographical group. For example

 

The last step is to create a new geographic group or select an existing geographic group. Here we select none.

 

Click Finish. Azure will automatically help you create this machine. At this time, the Creation status will appear below. After about 10 minutes, the machine will be automatically created.

 

After the creation is complete, select this machine and there will be various operation buttons, such as connection, start, close, attach new disk, capture image, delete and so on. For the connection, start, and add new disks, we will discuss them later, and capture images will be described in later sections.

 

Machine

 

First, the first button is connection. When you click Connect, an RDP file will be automatically downloaded to connect to the server remotely. The difference is that the password has been set manually, therefore, you do not need to use a pem file like amaozn to decrypt the private key. The machine name is hotcanazuresample.cloudapp.net. Azure does not set a DNS name through a public IP address, but allows users to select a DNS name for access. This name will not change, you do not need to bind elastic IP to this server. If you need to use your own domain name, you can directly use a cname record to link to this custom global DNS.

We can see that there is also a public IP address on the right. This public IP address may also change, so do not bind the record of your domain name to this IP address.

After connecting to the machine, let's take a look at the hardware configuration of the machine:

 

Similarly, this machine is amd opteron processor memory 1he, 2.1 GHz, GB memory. Unlike AWS, AWS has a 30 GB system disk and a GB temporary storage disk, which can be used to store temporary data. You must note that the data on this disk is not persistent. Therefore, if you need to save the data on this disk permanently, You need to mount a new disk instead of storing the data on this D disk.

Interestingly, this machine also carries a floppy disk and an optical drive, but I am not very clear about how to use this software and optical drive. Do you have to rush to the azure data center to plug the CD? I wonder if azure supports this service.

 

Create new volume

To make the system have persistent storage, We can click the attach button on the management interface to mount a new disk to Windows.

 

First, the vhd file will be automatically placed under the same storage account as the system disk of the machine, and the file name will be automatically generated. The size can be 1-GB. The last host cache defines the cache features of this disk, including no cache, read-only cache, or read/write cache. The default value is none. Note that if the read/write cache is used, data on the disk may be lost. This is because if the data is only written to the disk cache but not actually written to the disk, if a problem occurs on the physical machine where the machine is located and the machine needs to be upgraded or shut down, data that has not been written will be lost. However, if you never use the cache, performance problems may occur.

AWS is doing better at this point. AWS supports setting iops to solve disk performance problems and allocates disk read/write to different storage devices, azure does not support it yet.

After a disk is created, a new disk is displayed in windows. You can use it directly. Here, AWS is slightly different from AWS. When a disk is created, AWS is offline, while Azure is online.

 

 

Because the f disk is stored in the storage account, the data is durable. All required data can be stored on the F disk.

Endpoint)

When creating the azure Vm, I don't know if you noticed that we didn't directly configure the firewall in the Wizard, because this step is not directly included in the azure VM creation wizard, however, we still need to create an endpoint for this machine to allow external machines to access it. By default, azure also creates a port 3389 for remote connection for Windows.

Click the endpoint label to add a new endpoint.

 

Click Add endpoint:

 

Here there are two main options: one is a normal endpoint and the other is the Server Load balancer endpoint. We will explain it later. First, select Add endpoint. Enter the name, select the protocol, and define the external port and internal port. We should open port 80, so select TCP, both inside and outside are 80.

 

Click Finish. After the creation is complete, a port 80 is available for access. Note that the endpoint here is equivalent to the port ing on the vro. Sometimes you need to enable the firewall on the Windows machine.

 

Server Load balancer endpoint

Unlike AWS, azure uses something called the load balancing endpoint to achieve load balancing. But first, we need to create a new machine in the same cloud service. When creating a VM, select connect to an existing VM.

 

Connect to the machine you just created. After the machine is created, add a new endpoint to the endpoint tab. This time we chose the load balancing endpoint, select the http-80 just now.

 

Give this endpoint another name: http2, internal port or 80. We can see that there is only one external port for the load balance endpoint, and different ports can be configured internally. However, this is just a very simple Server Load balancer. Compared with the configurations provided by AWS, this Server Load balancer is very simple.

 

Note that there is a small bug in the management portal. The Server Load balancer name should not be the same as the Server Load balancer name. Otherwise, the Server Load balancer cannot be created and no error is reported.

Open the endpoint attribute and we can see that this is an endpoint for load balance. There are two virtual machines, including the two machines I just created. On the overall interface, load-balanced = yes is also displayed.

 

 

The configuration for creating virtual machines on Azure is basically the same as that on AWS, And the Wizard is simpler and easier to use. However, the functions provided, especially load balance, are simple to use and configure. It is far from enough to configure advanced LB for general use. Of course, the azure virtual machine is still in the preview stage. We hope that the LB function will be enhanced when it is officially released.