OpenStack platform summary (OpenStack platform)

OpenStack platform

My name is Anatoly Vasilenko and today I would like to tell you about OpenStack software platform. However, before I am going to introduce it to you I will tell about Cloud Computing.

Cloud computing is the division of labour concept in IT scope. It is very popular for already more than 5 years already. There are three main types of services which can be served to the customer: Infrastructure as a Service, Platform as a Service and Software as a Service. They are forming stack and varies in abstraction level offering to the customer virtual machines, or computing system or the software. Because of the division of labour clouds can gain benefits like reduction of the overall costs for resources, electric power and other datacenter costs, and at the same time get better scalability, redundancy, availability, productivity and other business significant benefits.

OpenStack is a free and open-source cloud-computing software platform. It was created about 5 years ago and now has a big community of developers, users and vendors around it. Even has its own governance. The mission of the OpenStack is to create universal platform for managing clouds of different purposes.

However, the OpenStack is not the only one cloud software platform, but it is the most upcoming or at least there is more marketing about OpenStack comparing to the other platforms from Amazon or OpenSwift or others. Between other existing systems, Amazon cloud is the most similar to the OpenStack. OpenStack even has the same API, enabling the applications written for Amazon to be freely ported onto the OpenStack. This gives a huge lift up on the path of conquering the world, because Amazon has the biggest cloud platform nowadays.

OpenStack is only five years old, but since its creation different companies started to maintain this project and the community started to grow up as a snowball.

OpenStack can manage different operating systems – Linux, windows or bare metal, and can cooperate with different types of hypervisors, such as Xen, VMware, KVM or even Linux containers for those who need minimal virtualization overhead.

OpenStack has modular architecture, what makes it more flexible and enables separate development of different component and services. Good API specification makes components even more independent. Now user, who is setting up its own cloud, can use only those components which are required by his applications.

There are three type of services: core services, that are needed in every cloud setup, shared services, that are the most widespread services giving additional functionality and other services, that are not commonly used, such as Designate service, which provides DNS as a Service.

Now, I would like to introduce to you some of the OpenStack components. There are 16 of them:

1. Nova is a compute service; it is a cloud fabric controller that is designed to manage pools of computer resources.

2. Glance is an image service, which manages the operating system images, it can store, backup and even migrate them while running.

3. There is two types of storages available in the OpenStack: Object storage called Swift, which purpose is to store and retrieve data as objects with simple API, and Block storage called Cinder, which is used to attach emulated block devices to virtual machines.

4. Horizon is a dashboard, providing administrators and users with the graphical interface to access and manage their cloud.

5. Keystone is an identity service, which is used to authenticate users and grant to them some access permissions.

6. Neutron is an OpenStack networking controller, it manages the network, controlling services like DHCP, static IP addresses, firewall, load balancer, virtual private network and some other network services.

7. Heat is an orchestrator. Orchestration is the concept of unified management and exception handling. Heat can set up your whole cloud by using only template with the description of the infrastructure. The templating language is also compatible with configuration management tools such as Puppet and Chef.

8. Ceilometer is a telemetry service, which is capable of collecting usage and performance statistics across the services. This is one of the most needed service for hosting providers, because it establishes the customer billing.

9. Trove is a database service focused on automation of complex administration task.

10. Sahara is the OpenStack realization of Hadoop.

11. Ironic is special project, which allows managing not the virtual machines, but running the bare metal. It is based on PXE and IPMI interface technologies.

12. Zaqar is the OpenStack realization of the queue service, it is based on the principles of Amazone similar project, but additionally handles broadcasting.

13. Manila is the realization of shared file system.

14. Designate is a service introducing DNS as a service.

15. And Barbican is a security project, which main goal is to create a centralized secret-store for keying material. This project is targeted to create its own community and to become not only the part of the OpenStack project, but to become a standalone project in our world.

In conclusion, I would like to say that OpenStack has well thought-out architecture and it seems that it achieved its main goal and now rapidly spread across the market.