[Report]Selecting and optimising Amazon EC2 instances #AWSSummit

2022.06.15

この記事は公開されてから1年以上経過しています。情報が古い可能性がありますので、ご注意ください。

This time, I will report on the session of AWS Summit Online India 2022, which is being held for two days from May 25 to 26, 2022. Use it to understand the session summary and check the sessions you are interested in. Also, the session archive will be published, so please check there for details.

This article is a report article on  Selecting and optimising Amazon EC2 instances

Session overview

Summary: Amazon EC2 offers the broadest and deepest platform to run workloads ranging from simple web applications to mission-critical workloads. There are now over 500 Amazon EC2 instance types to choose from to achieve the best price-performance for your cloud workloads.

Join this session to learn what tools you can use to select the right instance types. Also learn how to grow your compute capacity while rightsizing it by using Amazon EC2 Compute Optimizer and optimising costs using a mix of Amazon EC2 Spot, savings plans and on-demand instances.

Session Level

100 - Introductory

Agenda

Amazon EC2 instance options Selecting the right EC2 instance types Optimising EC2 instances for your workload • Price optimization - Optimal Purchase Options • Capacity Optimization - Auto Scaling groups • Performance Optimization – AWS Compute Optimizer

How do you select the right Amazon EC2 instance types for your workload?

Tip #1

Familiarize yourself with instance-type categories but no need to memorize everything

General-purpose Burstable Compute intensive Memory intensive Storage (high I/O) Dense storage GPU compute Graphics intensive

Tip #2

Focus on using Amazon EC2 efficiently Don’t try to completely optimize for cost on initial instance type selection Optimize with purchasing options and tools as you learn more about your consumption patterns

Tip #3

There is a good chance multiple instance types can power your application/workload Think flexible!

Tip #4

Take an attribute-based approach to instance type selection Work backwards from your application/workload performance requirements

Attribute-based approach for instance Selection

Focus on the capabilities and options that are required for your application/workload

What attributes do you need?

Choice of processor (AWS, Intel, AMD)

Fast processors (up to 4.0 GHz)

High mem footprint (up to 24 memory TiB)

Instance storage (HDD and NVMe)

Accelerated computing (GPUs and FPGA, Amazon Elastic Inference)

Networking (up to 100 Gbps, Elastic Fabric Adapter)

Bare metal Size (Nano to 32xlarge)

Attribute-based instance type selection (ABS)

Helps discover and select instance types to be launched by EC2 Fleet and Amazon EC2 Auto Scaling groups based on your required attributes

• Express instance requirements as a set of attributes, such as vCPU, memory, storage type, and much more!

• Ideal for instance-type-flexible workloads and frameworks, such as containers, big data, and CI/CD

• Future-proofs fleet configuration by automatically adding newly released instance types

• Enables you to protect the price

Optimizing Amazon EC2 cost and capacity

Pricing

Achieve optimal price and performance with different purchase models

  • Use Savings Plans for known/steady-state workloads
  • Scale using On-Demand for new or stateful spiky workloads
  • Scale using Spot Instances for flexible, fault-tolerant workloads

Capacity

Capacity management is made easy on the broadest and deepest compute platform

Using Amazon EC2 Auto Scaling

Performance

Enable AWS resource recommendations to reduce cost and improve performance

  • Lowers costs and improves workload performance
  • Applies insights from millions of workloads to make recommendations
  • Saves time comparing and selecting optimal resources for your workload

summary

Amazon EC2 is designed to provide the right computer for every workload

Take an attribute-based approach to instance type selection using tools like Attribute-based Instance Selection and Amazon EC2 Instance Selector

Leverage Spot and Savings Plans to acquire compute at a lower cost and innovate faster. Automate cost and capacity optimization

Optimize workloads by leveraging best practices via AWS Compute Optimizer

At the end

got to know tips, ways and Best practices to select ec2 Instance for saving money. Be sure to sign up for AWS Summit Online and check out the session archives! Slide materials were also distributed, so check it out.