AWS Chapter 2 Flashcards
Amazon EC2 Instance
a cloud instance to run/host applications/workstations/etc
EC2 Instance Benefits
launch within minutes, pay for only the compute time you use when running, pay for only capacity that you need
Types of EC2 Instances
General Purpose, Compute Optimized, Memory Optimize, Accelerated Computing, Storage Optimized
General Purpose EC2 Instance
Balance of all resources, used for app servers, small/medium databases
Compute Optimize EC2 Instance
Ideal for compute bound applications, high performance web servers etc
Memory Optimized EC2 Instance
ideal for fast performance for workloads that process large datasets
Accelerated Computing EC2 Instance
use hardware accelerators to perform functions efficiently, graphic processing
Storage Optimized EC2 Instance
workloads that require high, sequential read and write access to large datasets
On Demand Pricing EC2 Instance
ideal for short term, irregular workloads, only pay for what you use
Amazon EC2 Savings Plan
commit to a consistent amount of compute usage, 1 year/3 year terms, used with Cost Explorer
Reserved Instances
1 year/3 year term, at the end of the reserved instance time length you are charged On Demand pricing
Spot Instances
Ideal for workloads with flexible start/end times, don’t require contracts
Dedicated Hosts
physical servers with EC2 instance capacity that are fully dedicated to your use, most expensive
Dynamic Scaling
scaling of an EC2 instance that responds to changing demand, only pay for what you use
Predictive Scaling
schedule the right number of EC2 instances based on predicted demand
Monolithic Applications vs Microservices
Monolithic consists of an application that includes tightly coupled components, if one fails the rest fail. Micro are loosely coupled and if one fails it can continue
Amazon Simple Notification Services (Amazon SNS)
publish/subscribe service, can be email addresses, web servers, etc
Amazon Simple Queue Service (Amazon SQS)
message queuing service, can store send and receive messages between software components, messages are processed and then deleted
Serverless Computing
code runs on servers, do not need to maintain those servers, different from EC2 instances
Amazon Lambda
AWS serverless computing, only pay for compute time/when code is running. to run, upload code, send trigger, code runs
Amazon Elastic Container Service (Amazon ECS)
highly scalable container management system, supports Docker, packages code into a single object
Amazon Elastic Kubernetes Service (Amazon EKS)
runs Kubernetes on AWS, open source software that enables you to deploy container apps at scale
AWS Fargate
serverless compute engine for containers, manages server infrastructure for you
