AWS Billing and Pricing Flashcards
AWS Pricing Model
AWS works on a pay as you go model in which you only pay for what you use, when you are using it.
If you turn off resources, you don’t pay for them (you may pay for consumed storage).
There are no upfront charges, and you stop paying for a service when you stop using it.
Aside from EC2 reserved instances you are not locked into long term contracts and can terminate whenever you choose to.
Volume discounts are available so the more you use a service the cheaper it gets (per unit used).
There are no termination fees.
3 Fundamental Drivers of Cost with AWS
compute,
storage,
outbound data transfer.
Data transfer charges
In most cases, there is no charge for inbound data transfer or for data transfer between other AWS services within the same region (there are some exceptions).
Outbound data transfer is aggregated across services and then charged at the outbound data transfer rate.
AWS Free Tier
Free tier allows you to run certain resources for free.
Free tier includes offers that expire after 12 months and offers that never expire.
Pricing policies include:
Pay as you go.
Pay less when you reserve.
Pay even less per unit when using more.
Pay even less as AWS grows.
Custom pricing (enterprise customers only).
Free services include:
Amazon VPC.
Elastic Beanstalk (but not the resources created).
CloudFormation (but not the resources created).
Identity Access Management (IAM).
Auto Scaling (but not the resources created).
OpsWorks.
Consolidated Billing.
Amazon EC2 pricing is based on:
Clock hours of server uptime.
Instance configuration.
Instance type.
Number of instances.
Load balancing.
Detailed monitoring.
Auto Scaling (resources created).
Elastic IP addresses (charged if allocated but not used).
Operating systems and software packages.
There are several pricing model for AWS services, these include:
On Demand
Dedicated Hosts
Dedicated Instances
Spot Instances
Savings Plan
Reserved Instances
On Demand EC2 Pricing Model
Means you pay for compute or database capacity with no long-term commitments of upfront payments.
You pay for the computer capacity per hour or per second (Linux only, and applies to On-Demand, Reserved and Spot instances).
Recommended for users who prefer low cost and flexibility without upfront payment or long-term commitments.
Good for applications with short-term, spiky, or unpredictable workloads that cannot be interrupted.
Dedicated Hosts EC2 Pricing Model:
A dedicated host is an EC2 servers dedicated to a single customer.
Runs in your VPC.
Good for when you want to leverage existing server-bound software licenses such as Windows Server, SQL Server, and SUSE Linux Enterprise Server.
Also good for meeting compliance requirements.
Dedicated Instances EC2 Pricing Model:
Dedicated Instances are Amazon EC2 instances that run in a VPC on hardware that’s dedicated to a single customer.
Dedicated instances are physically isolated at the host hardware level from instances that belong to other AWS accounts.
Dedicated instances may share hardware with other instances from the same AWS account that are not Dedicated instances.
Spot Instances EC2 Pricing Model
Purchase spare computing capacity with no upfront commitment at discounted hourly rates.
Provides up to 90% off the On-Demand price.
Recommended for applications that have flexible start and end times, applications that are only feasible at very low compute prices, and users with urgent computing needs for a lot of additional capacity.
In the old model Spot instances were terminated because of higher competing bids, in the new model this does not happen, but instances still may be terminated (with a 2-minute warning) when EC2 needs the capacity back – note: the exam may not be updated to reflect this yet.
Savings Plans Pricing Model
Commitment to a consistent amount of usage (EC2 + Fargate + Lambda); Pay by $/hour; 1 or 3-year commitment.
Reservations Pricing Model:
Reserved instances provide significant discounts, up to 75% compared to On-Demand pricing, by paying for capacity ahead of time.
Provide a capacity reservation when applied to a specific Availability Zone.
Good for applications that have predictable usage, that need reserved capacity, and for customers who can commit to a 1 or 3-year term.
Reservation options include no upfront, partial upfront and all upfront.
Reservation terms are 1 or 3 years.
Reservations apply to various services, including:
Amazon EC2 Reserved Instances.
Amazon DynamoDB Reserved Capacity.
Amazon ElastiCache Reserved Nodes.
Amazon RDS Reserved Instances.
Amazon RedShift Reserved Instances.
Amazon Simple Storage Service (S3) Pricing
Storage pricing is determined by:
Storage class – e.g., Standard or IA.
Storage quantity – data volume stored in your buckets on a per GB basis.
Number of requests – the number and type of requests, e.g., GET, PUT, POST, LIST, COPY.
Lifecycle transitions requests – moving data between storage classes.
Data transfer – data transferred out of an S3 region is charged.
Amazon Glacier pricing
Extremely low cost and you pay only for what you need with no commitments of upfront fees.
Charged for requests and data transferred out of Glacier.
“Amazon Glacier Select” pricing allows queries to run directly on data stored on Glacier without having to retrieve the archive. Priced on amount of data scanned, returned, and number of requests initiated.
AWS Snowball Pricing
Pay a service fee per data transfer job and the cost of shipping the appliance.
Each job allows use of Snowball appliance for 10 days onsite for free.
Data transfer in to AWS is free and outbound is charged (per region pricing).