Pat Shuff's Blog

Oracle Linux on Oracle Compute Cloud

In this blog we are going to look at creation of an Infrastructure as a Service foundation using Compute as a Service and Storage as a Service to create an Oracle Linux instance. To start with we begin by logging into http://cloud.oracle.com and entering our Identity Domain, user name, and password. In this example we are connecting to metcsgse00026 as cloud.admin for the username.

If we look to the right of the Compute Cloud Service header we see a "Service Console" tab. Clicking on this allows us to create a new virtual machine by clicking on the "Create Instance" button. Not all accounts will have the create instance button. Your account needs to have the funding behind generic compute and the ability to consume either metered or un-metered IaaS options.

Note that we have two virtual machines that have previously been created. The first listed is a database service that was created. The compute infrastructure is automatically created when we create a database as a service instance. The second listed is a Java service that was created through the Java Service console. The compute infrastructure was also created for the JaaS option. We can drill into these compute services to look at security, networking, and ip addresses assigned.

To create a virtual machine we click on the "Create Instance" button which takes us to the next screen. On this screen we enter the name of the virtual machine that we are creating, a description label, the operating system instance and type, the shape of the instance. By shape we mean the number of processors and memory since this is how compute shapes are priced.

To select the different types of operating systems, we can enter a "*" into the Image type and it lists a pull down of operating system types. You can scroll down to select different configurations and instances. In the screen shot below we see that we are selecting OEL 6.4 with a 5 GB root directory. The majority of the images are generic Linux instances with different disk configurations, different software packages installed, and different OS configurations.

The next step is to select the processor and memory size. Again, this is a pull down menu with a pre-configured set of options. We can select from 1, 2, 4, 8, and 16 virtual processors and either 15 GB of RAM or 30 GB of RAM per processor. These options might be a bit limiting for some applications or operations but are optimized and configured for database and java servers.

In this example we selected a 1 virtual processor, 15 GB of RAM, 5 GB of disk for the operating system, and Oracle Linux 6.4 as the operating system. We can enter tags so that we can associate this virtual machine with a target, production environment, system, or geographic location consuming the resources.

At this time we are not selecting any custom attributes and not using Orchestration to provision services, user accounts, passwords, or other services into our virtual machine. We click the "Next" button at the top of the screen to go to network configurations.

In the network configuration we can accept the defaults and have an ip address assigned to us. If we have an ip address on reserve we can consume that reserved address and even assign a name to it to resolve to linux6.mydomain.net if we wanted to map this to an internet name. In this example we just accept the defaults and try not to get too fancy on our first try. This will create an ip address for our server, open port 22 for ssh access, and allow us to network it to other cloud services inside our instance domain with local network routing.

The next step is to configure a disk to boot from. We are presented with the option of using a pre-configured existing disk or creating a new one. The list that we see here is a list of disks for the database and java servers that we previously created. We are going to click on the create new check box and have the system create the disk for us.

The storage property pull down allows us to select the type of disk that we are going to use. If we are trying to boot from this disk, we should select the default option. If we were installing a database we would select something like the iSCSI option to attach as the data or fast recovery disk options.

The final step is to upload the public key of our ssh key pair. This is probably the biggest differential between the three services. Amazon requires that you use their shared and secret key that they generate for you. Microsoft allows you to create a service without an ssh key and use a username and password to connect to the virtual machine. Oracle requires that you use the ssh public-private key that you generate with puttygen or ssh-keygen. The public key is uploaded during this creation time (or selected if you have previously uploaded the key). The private key is presented via putty or ssh connection to the server once it is created. The default users that are created in the Oracle instances are the oracle account that is part of the orainst group and the opc account that has sudo privileges.

Once we have everything entered, we click on next and review the screen. Clicking on the "Create" button will create a compute instance with the configuration and specifications that we requested. In this example we are creating a Linix 6.4 instance onto a 1 OCPU machine with 15 GB of memory and attaching a 5 GB disk drive for the operating system.

As the system is provisioning, it updates us on progress

When everything is done we can view the details of the virtual machine and see the ip address, which key was used, and how the service is configured.

Before we can attach to the server, we need to open up the ssh port (port 22). This is done by going into the Network tab and adding a "Security Rule". This rule needs to be defined as public internet access to our default security rule since we accepted the default network rules as our network protocol in the creation of the virtual machine.

Note in this example we selected ssh as the protocol, public internet as the source, and default as the destination. With this we can now connect using putty from our Windows desktop. We need to configure putty with the ip address of our virtual machine and our private key as the ssh protocol for connecting. We could create a tunnel for other services but in this example we are just trying to get a command line login to the operating system.

Note that we can confirm that we are running Linux 6.4 and have a 5 GB hard drive for the operating system.
The whole process takes a few minutes. This is relatively fast and can be scripted with curl commands. More on that later.

Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.