With the emergence of the cloud services, everybody agrees that multiple roles within IT departments will change. Some of them will be changed more or less radically. Others will be completely removed.
Before the cloud, all the resources that were managed by IT were on-site. This means that the physical servers were in the organization’s datacenters (collocation is an exception where the servers are placed in 3rd party datacenters). The IT department was installing the software on those servers and was managing the software by updating it, patching it and using it.
Let’s think that a brand new company needs for all employees to have email addresses. How can IT take care of this?
First, they would have to acquire the server where the email server application will be installed.
Then that server needs to be racked, cabled and powered on in the datacenter.
Then the software will have to be installed and configured.
Lastly the email accounts will be created and the server will be monitored for failures, disk usage issues, CPU loads and so on.
Before we go ahead, I think we should refresh the three main cloud services offerings available right now:
* IaaS – Infrastructure as a Service – virtualized resources such as CPU, storage and network access. This is like deploying a physical server. You can do whatever you want with this virtual machine.
* SaaS – Software as a Service – provides access to traditional applications, word processors being one of them. The users have the same features like the ones that they had installed on their PCs.
* PaaS – Platform as a Service – used to develop applications with the tools/environments of the cloud provider.
Now considering IaaS and SaaS services and the case study from the beginning of the article.
If a customer would go with IaaS he/she will never have to deal with buying the server, cabling, power nor what is happening when the server will crash.
These services are provided by the IaaS provider through a virtual machine (VM).
However, the user will still need to install the software on the VM, configure it, add the email account and monitor it for CPU load and disk usage. In case he/she needs more of any of these, there is always the option to add more.
If the customer chooses SaaS, then the only thing that he/she will need to do is to configure email accounts through the application provided by the cloud services vendor. The application is monitored by the cloud provided, so monitoring is not needed.
If one would choose IaaS, then not much of the current job will change. The engineer will be relieved from physical work. But most of the job of the IT administrator will be the same.
However, there are new skills that need to be learned such as automation, virtualization and service management.
Let’s see how simple it is to deploy servers in Amazon Web Services (AWS).
AWS is well known for its IaaS offerings.
This will show you how you can deploy a VM in minutes. In this case, I will use an Amazon Machine Image, which is the Amazon Linux VM.
Once you login, you will be redirected to the AWS Management Console:
Click on EC2 to get to the EC2 Dashboard:
Click on Instances to create an instance and then on Launch Instance to start the process.
From the list, choose Amazon Linux AMI and click on Select to go to the next step:
Choose Micro Instances and then click on Next: Configure Instance Details
Leave everything as it is on Step 3 and click on Next: Add Storage:
8GB will be enough and continue by clicking on Next: Tag Instance:
Tag your instance by configuring a pair of key/value and go to Next: Configure Security Group:
Leave this step as it is and click on Review and Launch to get an overview of all the settings:
Click on Launch to continue:
Click on acknowledgement and then on Launch Instances:
To view the instances, click on View Instances:
As you can see, we have only one instance.
To login to the instances, you will need to get the full value of Public DNS column:
If you are using Windows to connect to the instance, download putty, add the .ppk file saved earlier and in Host Name field type firstname.lastname@example.org (the string after @ will be different in your case based on where you choose to deploy your VM, the IP address).
If you configured correctly putty, you should be able to connect to the VM:
From now on, a server administrator is almost on its own. The software has to be installed and configured and all the other monitoring jobs specific to the software installed have to be performed continuously.
As I said, more or less, some job roles will be changed to adapt more for the cloud.
But all these skills need to be developed by everyone:
* In-depth technical skills regarding Internet technologies – as everything will be done over the Internet with a focus on frameworks like Java, .NET among others.
* Understand how the cloud is working – and here is about to understand virtualization. As one might already figure it this out, everything in the cloud is virtual. The good part is that you most likely have run into virtualization already, even if you installed one single VM in VMware. Also the immediate cost of the cloud has to be understood as well. Keep in mind that regardless of whether you use or don’t have a resource that’s allocated to you, you are still paying for it.
* Negotiation – you should know how to negotiate with the cloud providers.
* Security – you should be aware of how the data is transmitted securely as now it will go out of your inhouse servers and be stored in the cloud. Also you should get familiar with whom and how they can handle your data based on the country where you choose to deploy the data in the cloud.
This article is showing how you can deploy quickly a server in Amazon AWS cloud and briefly show what will be the new skills that IT guys will need when they make the transition to cloud.