Descriptive part
Cloud and autonomic computing is a subject that allows us to discover the use of the cloud around three main axes: via virtual machines such as Virtual Box, Docker and OpenStack.
These three tools are not based on the same thing. Indeed, we must differentiate between virtual machines and containers. To simplify things, a virtual machine is based on a hypervisor which runs on a host OS and which allows to virtualize one or more guest OS. There are two types of hypervisors. Type 1 where the hypervisor runs directly on the hardware and type 2 where it runs on the kernel. Instead of virtualizing the hardware, a container run multiple virtual instances of the same OS over a single hardware. It provides isolation between processes and appear as a separate OS.
A container is much lighter and faster than a virtual machine. It only needs a few seconds to start up, compared to a few minutes for a virtual machine. Finally, it takes only a few MB, which is 50 to 100% less than a VM. However, there are also differences depending on the type of container. For more information, I invite you to download and read my report here.
Virtual Box
VirtualBox is a type 2 hypervisor, meaning that it must be installed on an operating system, not directly on a computer as an operating system.
The VM’address is a subnetwork of the host (mask 255.255.0.0 for the host and 255.255.255.0
for the VM). We can now try the interconnectivity.
The VM has accessed to the outside web but is not directly visible from the host → VM is
isolated from the host. We can only reach the NAT network (192.168.56.1 here). Each VM
created in parallel will have the same IP Address (10.1.5.72).
Now, the VM can be accessed through the port 2222. We have the same machine with the same IP Address. Everything is shared (it’s normal) except the network. We need to open a new port on the host to access the VM.
Docker allows you to embed an application in one or more software containers that can run on any machine server, whether physical or virtual. Docker works on Linux as well as Windows Server. It is a technology that aims to facilitate application deployments, and the management of the underlying infrastructure sizing.
Docker uses a client-server architecture. The Docker client talks to the Docker daemon, which does the heavy lifting of building, running, and distributing your Docker containers. The Docker client and daemon can run on the same system, or you can connect a Docker client to a remote Docker daemon. The Docker client and daemon communicate using a REST API, over UNIX sockets or a network interface. Another Docker client is Docker Compose, that lets you work with applications consisting of a set of containers. The network is bridged meaning that it is shared within the containers.
CT2 is a copy of ubuntu that has been configured with nano installed. By using CT2, you can use nano without having to reinstall it because it is already there, unlike CT4 which was created from scratch.
One of the main advantages of Docker is the use of dockerfile which are configuration files. It allows to automate tasks which is useful when you have more than one container to launch.
Docker
Openstack
OpenStack is a platform for creating and managing private or public clouds based on pools of virtual resources. The tools (or "projects") that make up the OpenStack platform provide the main services of cloud computing, namely, computing, networking, storage, identity management and image management.
OpenStack leverages application programming interfaces (APIs) to push the boundaries of abstraction of these virtual resources by distributing them into individual pools, which drive standard cloud tools that administrators and users interact with directly.
Here is a general view of the network created with a virtual machine on its own private network. A router makes the connection between the private network and the outside world. However, in order to really access the VM from outside, we have to create a floating IP.
The next step is to deploy a 2-tier Web application on OpenStack. The application implements a calculator that handles arithmetic operations (i.e. addition, subtraction, multiplication and division) of integer numbers. Each one of these operations is implemented with an NodeJs micro-service. In addition, there is a fifth micro-service that implement the application endpoint. Specifically, this service will be listening and receiving the prospective requests when starting the application. Requests are HTTP-based.
Finally, we have the following architecture with the 4 operations on a separate private network and the Calculator Service (CS) on its own private network. It is the only one to have a floating IP. The two subnets are connected by a router.
Using a dockerfile, we can easily automate the creation of sub-services in the following way.
Once the route has been added between PN and PN2, we can start the CS microservice and check that it runs without any problem.