PA200 - Cloud Computing
Lecture 11: Future directions in cloud computing
by Ilya Etingof, Red Hat
Warm-up
Let's rehearse on the previous lectures...
On-premises applications
• Monolithic
• Tied to the infrastructure
• Languages: a Visual Studio language, enterprise Java, Cobol
• Developed in a waterfall model
Cloud-native applications
• Modular and stateless
• Shared resources
• Elastic and redundant by design
• Web-service architecture
• Rolling updates
• Agile, DevOps, CI/CI
Microservice architecture
Linux containers
• Form of concurrency and isolation
• Based on kernel features
• Control groups
• Namespaces
• Container orchestration
Container orchestration software
• Makes containers usable
• Docker - single node containers
• Builds container image (file system)
• Runs image as a container instance
• Kubernetes - multinode container cloud
• Extends Docker over multiple hosts
• Introduces cloud features
In this lecture...
• Future directions and challenges in cloud computing
• Course summary and walk through
Future directions
• Internet of Everything, Edge/Fog computing
• Containers as a unit of software distribution and deployment
• Serverless & FaaS
Edge computing (1/3)
• Internet of Things / Internet of Everything
• Smart homes / smart cities
• Connected cars / public transport
• Health care
• ...
• Localized data
• High-bandwidth media
• AI/ML
Edge computing (2/3)
Stretching the cloud
• Move processing and storage units closer to consumers
• Automate cloud maintenance at the edge (e.g. bare metal)
• Protect larger attack surface
Edge computing (3/3)
Containers and PaaS (1/2)
• Container images for software distribution
• CSP unification on Kubernetes
• Containers on the bare metal
Containers and PaaS (2/2)
Serverless & FaaS (1/2)
• FaaS keeps gaining popularity
• Emerging serverless databases
Serverless (2/2)
Conclusions
• Long way towards concurrency and isolation
• Cloud is powering almost everything
• OpenStack for IaaS, Docker/Kubernetes for PaaS
• Cloud keeps evolving
Rehearsal
• Virtualization technologies
• Cloud providers
• Introduction to OpenStack
• Cloud administration and software development
• Cloud software architecture and containers
What's virtualization
• Form of computation concurrency and isolation
• Dates back to early mainframes
• Progressed through
• Batch jobs
• Multitasking
• Multithreading
• OS virtualization
• Containers
What's hypervisor
• Performs virtual machines lifecycle management
• Two types: native and hosted
• Technologies: full and para-virtualization
• Nested virtualization
CSP roles
• CSP can do IaaS, PaaS, SaaS
• IaaS CSP abstract away hardware
• PaaS CSP abstract away OS
• PaaS/Serverless CSP abstract away application runtime
• PaaS/FaaS CSP abstract away language runtime
• SaaS CSP offer managed applications
OpenStack basics
• OpenStack is a free and open source IaaS software
• Operates on VMs (instances)
• HEAT engine consumes HOT templates to build infrastructure
• Key projects
• Nova, Glance, Swift, Neutron, Heat, Cinder
Technology behind OpenStack
• Based on open protocols
• Consists of many loosely coupled services
• Each service typically contains
• Service engine
• REST API engine
• Persistent database
• RPC e.g. message queue
OpenStack pillars
• Open source
• Open community
• Open design
• Open development
Cloud-native applications
• Modular and stateless
• Shared resources
• Elastic and redundant by design
• Web-service architecture
• Rolling updates
• Agile, DevOps, CI/CI
Container orchestration
• Linux containers base on kernel features:
• Control groups
• Namespaces
• Docker builds, shares and runs container images on a host
• Kubernetes can base on Docker to create container cloud
Kubernetes basics
• Cluster
• master + nodes (on bare metal or VMs)
• nodes run pods
• Pods
• Pod contains one+ containers
• Application runs in its pod
• Controllers
• Pod management logistics (e.g. Deployment, StatefulSet)
• Services
• Represent application to the world
We are done
Thank you!