AWS Cloud Practitioner Flashcards
What are the 6 advantages of cloud computing?
Trade capital expense for variable expense
Benefit from massive economies of scale
Stop guessing about capacity
Increase speed and agility
Stop spending money running and maintaining DCs
Go global in minutes - near infinite scale
What are the types of cloud computing?
IAAS - manage server and OS, e.g. EC2
PAAS - someone else manages underlying OS and machine, e.g. Elastic beanstalk
SAAS - you only use the software and how to use it, e.g. gmail
What are the types of cloud computing deployments/
Public cloud - AWS, Azure, GCP
Hybrid cloud - mixture of public and private
Private cloud (on prem) - in your datacentre using Openstack or Vmware
What’s the difference between AZs, regions and edge locations?
Availability zone - data centre (potentially a few near each other) - each have redundant power, networking etc.
Region - a geographical area - consists of 2 or more availability zones
Edge locations - endpoints for AWS which are used for caching content (e.g. for CloudFront CDN). Currently there are over 150 edge locations
What are the 4 types of support plan?
Basic, Developer, Business and Enterprise
What are the IAM user access types?
Programmatic and console access
What are the 3 different ways to access the AWS platform?
Console
Programmatically
SDK
What is best practice when creating a root account?
Secure password, enable MFA, create individual IAM users
What is the structure of a policy?
JSON document that defines effect (allow/deny), resource and actions
What is S3?
Simple Storage Service
Object based, as opposed to block-based (which is more appropriate for OS files, DB engine etc.)
Objects are stored as key/value pairs: key of the object (name) and value is the bytes (data)
Data is spread across multiple devices and facilities
What is the maximum object size in S3?
Storage is unlimited (max object size of 5TiB)
Do bucket names have to be unique?
Bucket name must be unique as DNS entry is created for the bucket
What are the data consistency characteristics of S3?
Read after write consistency for PUTs of new objects
Eventual consistency for overwrite PUTs and DELETEs
What are the percentage guarantees of S3 for availability and durability?
Built for 99.99% availability
99.9% availability guaranteed
Durability guaranteed to 99.99999999999% (11x9s) - very unlikely you will lose any data
What are the storage classes of S3?
Standard - 99.99% availability and 99.99999999999% durability. Stored redundantly across multiple devices in multiple facilities - designed to sustain the loss of 2 facilities concurrently
IA (Infrequently accessed) - for data that is accessed less frequently, but requires rapid access when needed. Lower fee than S3, but changed a retrieval fee
One Zone IA - lower cost than IA, where multiple availability zone data resilience is not required
Intelligent Tiering - uses ML to automatically move data to most cost-effective access tier without performance impact or operational overhead
Glacier - for data archiving. Cheap, and retrieval time can be configured from minutes to hours
Glacier deep archive - lowest-cost storage class where a retrieval time of 12h is acceptable
Are buckets global or per region?
When you view buckets, you view them globally, but you can have buckets in individual regions
Is the storage class of S3 per-bucket or per-object/
When changing storage class, this is done per object in a bucket (I suppose this allows lifecycle rules to work easily)
What are the charges for S3?
Storage
Requests
Storage Management Pricing
Data Transfer Pricing
Transfer Acceleration (fast, easy and secure file transfers over long distances - takes advantage of CloudFront edge locations. Data arrives at edge location and is routed to Amazon S3 over an optimized network path. There is a tool that will tell you have much faster it is to various locations from where you are)
Cross Region Replication Pricing (automatic replication of objects from a primary bucket into a secondary bucket in another region)
What is the URL format for statically hosted S3 websites?
URL format: .s3-website-.amazonaws.com
Does S3 scale automatically?
Yes.
S3 will scale automatically with demand - great for static sites where there will be a large number of requests (e.g. movie preview)
What is CloudFront?
Cloudfront is a CDN (content delivery network). A CDN is a system of distributed servers (network) that deliver web content to a user based on the geographic locations of the user, the origin of the webpage, and a content delivery server.
What are the types of CloudFront distribution?
Web Distribution - typically used for websites
RTMP - used for media streaming (adobe flash media protocol)
What is a CloudFront edge?
Edge - location where content is cached (separate from AWS region or availability zone)
What is a CloudFront origin?
Origin - the origin of all files that the CDN will distribute (can be S3 bucket, EC2 instance, elastic load balancer or Route53)
What is a CloudFront distribution?
Distribution - name given to the CDN which consists of a collection of edge locations
Are edge locations readonly?
No, edge locations are not just readonly - you can write to them, too (e.g. S3 transfer acceleration)
Is there a charge for purging CloudFront caches?
Cached objects can be cleared, but there’s a charge for each purge request
What is EC2?
Elastic Compute Cloud.
Virtual servers in the cloud.
What are the pricing models for EC2?
On Demand
Reserved
Spot
Dedicated Host
What is on-demand EC2 pricing? What workloads is it suitable for?
Payed by the hour (or second for linux).
No up-front payment.
Great for apps with short-term, spiky or unpredictable workloads that cannot be interrupted (e.g. dev/test)
What is reserved EC2 pricing? What workloads is it suitable for?
provides you with a capacity reservation, but offers a significant discount on the hourly charge for an instance. Contract terms a 1 year or 3 year terms. More paid up front, more discount you get (max discount is if all is paid up-front).
Great for apps with steady state or predictable usage
What are the 3 types of reserved EC2 pricing models?
Standard reserved instances - offer upto 75% off on demand instances. More you pay up front and the longer the contract, the greater the discount. Issue is that you can’t change product family (e.g. go from high performance compute to high performance memory)
Convertible reserved instances - offer upto 54% off on-demand capability to change the attributes of the RI as long as the exchange results in the creation of reserved instances of equal or greater value
Scheduled reserved instances - available to launch within time windows you reserve. Allows you to match your capacity reservation to a predictable recurring schedule that requires a fraction of a day/week/month
What is spot EC2 pricing? What workloads is it suitable for?
Spot - enables you to bid whatever price you want for instance capacity - great savings if your apps have flexible start/end times (instance provisioned when spot price met, and will be lost when price exceeds your spot price)
Good for apps with flexible start/end times
Good for apps that are only feasible at very low compute prices, e.g. genomics that can be done at night
Good for users with urgent compute needs who need a large amount of additional capacity
Does a customer pay for an hour of EC2 spot instance if the instance is interrupted?
If you don’t get a full hour of compute because the price goes up, you won’t be charged for a partial hour.
But if you terminate the instance, you will be charged
What is dedicated host EC2 pricing? What workloads is it suitable for?
Dedicated hosts - physical dedicated servers - help reduce costs by allowing you to use your existing server-bound software licences (quite rare - often used when some software being used requires a dedicated host)
Useful for regulatory requirements that may not support multi-tenant virtualization
Great for licensing that doesn’t support multi-tenancy or cloud deployments
Can be purchased on-demand (hourly) or as a reservation for upto 70% off the on-demand price
What are the EC2 instance types?
Spell out FIGHT-DR-MC-PXZ
F - FPGA I - IOPS (I/O per second) G - Graphics H - High disk throughput T - Cheap general purpose (e.g. T2 Micro) D - Density R - RAM M - Main choice (general purpose) C - Compute P - Graphics (think pics) X - Extreme memory Z - Extreme memory and CPU
What is AWS EBS?
Elastic Block Store.
Used for VHDDs
What are the types of volume available on EBS?
SSD:
General purpose SSD (GP2) - balances price and performance
Provisioned IOPS SSD (IO1) - highest-performance SSD (low latency or high throughout workloads)
Magnetic:
Throughput Optimised HDD (ST1) - low cost HDD volume designed for frequently accessed, throughout-intensive workloads
Cold HDD (SC1) - lowest cost, designed for less frequently accessed workloads (file servers)
Magnetic - previous generation
Do EBS volumes need to be in the same AZ as an attached EC2 instance?
Yes
Do EBS volumes auto-replicate?
Yes
If you use a private key to SSH to an EC2 box, which user do you login with?
ec2-user
What are security groups?
Firewalls
Are roles more secure than id/key pairs? Why
Yes - don’t need to put credentials on the EC2 instance.
How long to EC2 role changes take to become effective?
Role changes are effective immediately
Is IAM per region?
No, it is universal
What is AWS ELB?
Elastic Load Balancer?
What are the 3 types of ELB?
Application load balancer - can make decisions based off application state (can see into layer 7)
Network load balancer - when you need ultra-high performance and static IP addresses
Classic - previous generation
What is the OSI model?
Open Systems Interconnection - defines 7 layers of computational communication.
1 is physical layer, 7 is application layer.
4 is transport layer.
Which DBs are available on RDS?
SQL Server Oracle MySQL Server PostgreSQL Aurora (invented by Amazon) - compatible with MySQL - 6 copies of DB across availability zones MariaDB
How many copies of data are created for Aurora?
6 across different AZs
What are the two key features of RDS?
Multi-AZ (for disaster recovery) - apps use DNS name, not IP addresses of specific instances
Read Replicas (for performance) - replicates primary DB to read replicas. They’re NOT for failover - without Multi-AZ your DB would be down. The read-replicas have their own DNS names so things can be scaled out. Can have upto 5 copies.
What is Amazon DynamoDB?
NoSQL DB as a service. Can scale automatically.
An questions around being able to auto-scale but the type of DB isn’t mentioned - go with DynamoDB
What is Amazon Redshift?
Fully managed, petabyte-scale data warehouse service in the cloud
How many GB in a Petabyte?
1,000,000 GBs
What are the four main differences of cloud computing?
IT assets become programmable resources - no more procurement and racking/stacking servers. TIME TO MARKET MUCH QUICKER
Global, available, and unlimited capacity - can deploy around the world, don’t need to arrange in other countries
Higher level managed services - e.g. ML - don’t need to hire an ML specialist
Security build in - firewalls, DDoS protection, can security audit quickly in test environments
What are the 11 design principles of AWS architecture best practice?
Scalability (up and out) Disposable resources instead of fixed servers Automation Loose Coupling Services, Not Servers Databases Managing Increasing Volumes of Data Removing single points of failure Optimize for cost Caching Security
What is AWS EMR?
Elastic MapReduce
What is AWS SQS?
Simple Queue Service
What is the minimum number of AZs in a region?
2
What is the AWS pricing philosophy?
AWS philosophy on pricing: pay for what you used at the end of each month, start/stop using a product at any time, no long term contracts required