Overview
Teaching: 10 min Exercises: 0 minQuestions
Why and when should we use the cloud?
How do we use the cloud?
Objectives
Understand cloud scalability and use in research
You do not wait for compute tasks to go through a queue
Compute can start as soon as you want it
You do not purchase and maintain hardware, operating systems etcetera
Upgrades just happen
You pay for resources you use; and then shut them off
You don’t have to buy into an institutional cluster if the cost calculation doesn’t make sense for you.
You have huge scale-up potential (reduced processing time)
In principle, you have near-infinite computing capacity.
There is a huge support community rapidly expanding cloud tools and tech
Because of the public availability of these resources, and substantial buy-in from industry, there is a large eco-system of tools and resources.
Storage, reliability, security and many other off-the-shelf services
And they just keep making new stuff.
You don’t have time to learn how to work on the public cloud There is stuff to learn. That’s what we’re here for! But there will be more to learn after this session is over. If you prefer to learn other things, you might not want to invest your time in learning about the cloud.
You operate your computer(s) at a very high duty cycle (more cost-effective) If your computer is constantly computing something, the cloud might end up costing you more.
Compute (=EC2)
Storage (= S3)
Manage = Databases… SQL, Not Only SQL… Data Warehouses… query machinery
Web = Web services, web sites, APIs, Clients, confederation, …
Services = All of the above simplified: Often no Compute involved
Key Points