Basics of Cloud Computing
-Notes by Saransh Kejriwal
Advantages of a cloud over having your own in-house setup are the same as those of hiring an Uber over owning a
car. Won’t dwell on that…
Cloud environments are classified based on 2 considerations – - Type of Service model
- Type of Deployment model
Deployment models – Public, Private, Hybrid:
Public – A bus, you share it with others, the provider owns it (eg. IBM BlueCloud, Amazon AWS, Microsoft
Azure)
Private – your own car (for an org only; huge upfront cost, eg. VMware, AWS)
Hybrid – Uber (effectively you’re using your local setup in tandem with the cloud to get the best of both
worlds)
Hybrid clouds are typically used to maintain confidential data privately.
Service models – Iaas, Paas, Saas
Iaas (Infra as a Service) – if all you need is a VM and you know how to setup your environment.
Paas (Platform as a Service) – a platform to create your own program
Saas (Software as a Service) – the finished program to use as an abstracted end user on the net. With a
customisable app.
Iaas – AWS EC2, Google, Azure; used by IT admins
Paas gives you a platform ready runtime envt. Eg. AWS EBS, Azure, Google App Engine; used by developers
Saas – vendor manages everything. We pay as we go.
Digital Ocean was the 3rd largest in client-facing hosting service. VMware is owned by Dell.
Using Cloud
Once you know why you need a cloud service, you need to choose the right hardware for your usage, eg:
Hardware services
EC2
Lambda – for “serverless computing (backend as a service, but you only pay for what you’re using
dynamically…somehow)”
ECS (Elastic Container Service)
Storage services – to backup and/or archive data
S3 – used for Backup
Glacier – used for Archival
EFS
Network Services
VPC – for network (???)
Route S3 – DNS
Direct Connect – P2P
Security Services
IAM – Auth
KMS - Data encryption
Monitoring tools
Cloudwatch – for monitoring
Autoscaling
CloudFormation – define the management process and tools (???)
Testing tools
Codestar
Codebuild
Codepipeline
Analytics tools
Athena – for visual representation
EMR – Elastic MapReduce
EC2 + S3
S3 can be used for data storage and EC2 can be used to run a linux envt.
Need a more comprehensive tute to get how to actually run anything on AWS.