OpenStack and Emerging Technologies

What it is

OpenStack is an open source cloud operating system. It controls large pools of compute, storage and network resources and offers them to users through APIs and a web dashboard, so users can create their own VMs, networks and disks on demand.

Emerging technologies such as containers, edge computing and serverless change how applications are packaged and where they run. The exam asks about them at concept level only.

How it works

OpenStack in brief

OpenStack was started in 2010 by Rackspace and NASA and is released under the Apache 2.0 license. A new release comes out about every six months.

OpenStack is not a hypervisor. It is a management and orchestration layer that drives hypervisors such as KVM, and storage and network systems, through drivers. Huawei builds its own cloud platform products on OpenStack.

The core projects

Two pairs cause most confusion. Cinder gives block volumes to instances, while Swift stores objects reached over HTTP. Ceilometer only collects data; evaluating alarms on that data is the job of a separate project, Aodh.

How the projects work together

Projects talk to each other through their REST APIs. Inside one project, components use a message queue, commonly RabbitMQ, and keep their state in a database.

Starting a VM touches almost every core project:

1. The user logs in and Keystone issues a token.
2. Nova checks the request, and its scheduler chooses a compute host.
3. The host gets the image from Glance, a network port from Neutron, and any volume from Cinder.
4. The hypervisor on the compute host starts the instance.

Containers

A container packages an application with its libraries and settings and runs it as an isolated process on the host OS. Containers share the host's kernel, while each VM runs its own full guest OS.

Docker made containers popular. An image is a read only template, a container is a running instance of an image, and a registry stores and shares images.

docker pull nginx
docker run -d -p 8080:80 nginx
docker ps

These commands download the nginx image, run it in the background with host port 8080 mapped to port 80 in the container, and list running containers.

Kubernetes

Kubernetes orchestrates containers across many servers. It schedules them, restarts failed ones and scales them up or down.

Edge computing and serverless

Edge computing processes data close to where it is produced, such as in a factory, a store or a base station, instead of sending everything to a central cloud. It lowers latency, saves bandwidth and keeps working when the link to the cloud is slow.

Serverless computing lets developers deploy code or functions while the provider runs and scales the servers. Functions run when an event triggers them, scale automatically, and are billed by actual execution, not by idle servers. FaaS (function as a service) is the most common serverless form.

A worked example

A retailer runs its online store on an OpenStack private cloud. Developers use Horizon to create instances; Keystone checks their identity, Nova places the instances, Neutron gives them networks and Cinder adds data volumes.

The web front end is moved into containers managed by Kubernetes, so it scales up for a sales event. Image resizing becomes a serverless function that runs only when a photo is uploaded. In each store, edge servers analyze camera video locally and send only the results to the cloud.

How the exam asks about this

• "Which OpenStack project provides the compute service?" -> Nova.
• "Which OpenStack project provides authentication and the service catalog?" -> Keystone.
• "Which statement about containers is true?" -> Containers share the host OS kernel, so they are lighter and start faster than VMs.