In this article, COMPAREX AG Senior IT Architect Eric Berg will talk about the use of virtual machines in the cloud, benefits and associated costs.Virtual machines are part of the Infrastructure As A Service solution in
Azure . By and large, there are no significant differences between the virtual systems in Azure and other systems that are hosted in the company's dedicated data storage and processing center. All of them have a specific set of hardware (CPU, RAM, HDD, network) and an operating system. However, the virtual machines used by Azure are strictly standardized to optimize resources and costs.
Examples of classes:
VM class | CPU cores | Ram | HDD # |
---|
A2 | 2 | 3.5 GB | four |
A11 | Intel Xeon E5-2670 16 cores @ 2.6 GHz | 112 GB | sixteen |
D4 | eight | 28 GB | 400 GB local SSD |
G5 | 32 | 448 GB | 64 6596 GB local SSD |
Expenses
Any definitive assumption that managing virtual machines in Azure will reduce costs is most likely erroneous. The number of costs arising from the use of VM in Azure is equal to the costs associated with the use of a special data center.
')
However, cloud VMs still have one significant advantage over their local receivers: costs arise only at the time of their actual use. In general terms, this is an additional advantage of virtual machines located in the cloud environment. For example, it is possible to minimize the number of servers in a remote desktop farm hosted in Azure at night, as well as use a “full” farm with its proportional daytime costs. Similarly, it will be quite possible to dynamically adjust VM performance and, as a result, actively respond to unsustainable functioning and resource requirements.
This kind of flexibility adds value to the virtual machines used in Azure. Moreover, when using virtual machines, the user may not use their own hardware.
Testing / Development
Performance testing or software development systems are the ideal scenario for using Azure VMs, since the initial benefits immediately manifest themselves even at a basic level.
The virtually unlimited resources means that even large development and testing environments can be equipped with reliable mobile devices.
Moreover, conducting a new test or allocating resources becomes available in Azure on demand. Cases of the use of internal processes, when testing and development were slow and unproductive, are already going down in history.
Service Level Agreement (SLA)
The issue of availability is key to allocating virtual machines in Azure. In order to shed light on this issue, it is important to understand how the data storage and processing centers are structured and where the VMs are located.
A key role in the operation of virtual machines in Azure is played by failure and update domains. In this case, failure domains describe systems with identical dependent objects, such as a power supply, network access, or server cabinets. Update domains are systems that use the same maintenance cycle, for example, the introduction of identical patches.
Do not forget that systems running in the same domain of failure and / or updates, under certain circumstances, may be subject to simultaneous failure!
That is why SLAs are not in any form for individual VMs in Azure. In order to get guaranteed availability, you must use at least two virtual machines in the same availability group. So the system will be configured in such a way as to guarantee the management of offering the same VM services on the leading computers. Thus, at least one virtual machine will always function in the availability group.
Service Level Agreement for virtual machines in Azure.Supply
There are several templates available in the Azure directory for delivering virtual machines. Templates are actually VM images that Microsoft regularly updates. These include
Windows Server operating systems, as well as Ubuntu or CentOS.
In addition to the delivery of virtual machines for catalog templates, there is also the possibility of working with images or VMs that users create independently.
First of all, it is possible to migrate 1: 1 virtual machines that were previously used locally in Azure. Here, the local VHD file is moved to an Azure storage account with which it can be launched using the Azure VM.
VM Migration to AzureThere is another method for storing VM images in a directory. Here, the virtual machine and all its settings can be prepared in a local environment. The system is prepared for deployment using the Sysprep utility, and after all configurations are completed, a suitable image is created. This image is then uploaded to Azure and used to create new VMs.
Upload your own image to AzureThe method of creating images in the cloud environment or for the mobility of a VM exists along with its hybrid counterparts. It provides for equipping the virtual machine with a suitable pre-configuration in Azure. Capturing VM is a method of moving this type of virtual machine to your image catalog, where it can be used to create new VMs.
Create your own image in AzureIn custody
Azure virtual machines that act as integral elements of an IaaS solution can be used flexibly, and the costs will be only during their actual use. This creates numerous scenarios in which modifications / deliveries can lead to a marked reduction in costs. However, all possible scenarios have one thing in common: hardware and other equipment can be ignored. As a result, users will focus on the really important thing - virtual machines.