saaQuestion
#1. S3 transfer fee
721p
s3๋ก์ ingress fee : $0
s3 to cloudfront : $0
s3 to internet, cloudfront to internet : ๊ธฐ๊ฐ๋ฐ์ดํธ๋น $0.1์์ค์ด๊ณ cloudfront๋ฅผ ํตํ egress๊ฐ ๋ ์ธ๋ค.
s3 transfer acceleration : $0.04 ~ $0.08 per GB
s3 cross region replication : $0.02 per GB
๊ฒฐ๋ก : S3 to internet์ด cost๊ฐ ๊ฝค ๋๋ค.
์ฃผ์ : storage cost(per GB per month) : ์ ์ฅ ์ฉ๋ ๋น์ฉ, retrieval cost๊ฐ ์กด์ฌํ๋ค.
#2. S3 versioning
Different versions of a single object can have different retention modes and periods : ๋ค๋ฅธ ๋ฒ์ ์ ์ค๋ธ์ ํธ๋ ๋ณด์ ๊ธฐ๊ฐ์ ๋ค๋ฅด๊ฒ ์ค์ ํ ์ ์๋ค.
object version์ ๋ช ์์ ์ผ๋ก Retain Until Date์ค์ ๊ฐ๋ฅ, (default ๊ธฐ๋ฅ ์๋)
๋ฒ์ผ์์ versioning์ ํ ๋ฒ ์ค์ ํ๋ฉด ๋ค์ unversioned๋ ์ ์์ง๋ง version์ suspendํ ์ ์์
#3. Resilient to periodic spikes in request rates : ๊ธ๊ฒฉํ ์์ฒญ ์ฆ๊ฐ์ ๋ํ ํ๋ณต ํ๋ ฅ์ฑ์ด ์กด์ฌํ๋๊ฐ
Cloudfront
regional failover์ ์ง์ํ๋ค๋ ์ ์์ resilientํ๋ค.
Aurora
read replica๋ ์ต๋ 15๊ฐ๊น์ง multi az in a region์ด ๊ฐ๋ฅํ๋ค
read replica๊ฐ write instance๊ฐ ๋ค์ด๋๋ฉด write instance๊ฐ ๋ ์ ์๋ค๋ ์ ์์ resilientํ๋ค.
#4. Storage Cost
EFS : $0.30 per GB
gp2(SSD) : $0.10 per GB
s3 : $0.023 per GB
s3 < gp2 < EFS ์์ผ๋ก ๋น์ธ์ง๋ง gp2๋ provisionํด์ผ ํด์ ์ฌ์ฉ๋ ๋งํผ ์ง๋ถ๋๋ ๊ฒ์ด ์๋๋ค.
#5. EBS volumes
io2 Block Express : 256,000IOPS
io2 : max 64,000IOPS for nitro(nitro๊ฐ ์๋ ๊ฒ์ ์ต๋ 32,000IOPS), size๋ 4GiB to 16TiB
io1/2๋ง multi-attach๊ฐ๋ฅ gp๋ ๋ถ๊ฐ๋ฅ
#6. SQS Batch
SQS FIFO Queue๋ ์ด๋น 300๊ฐ๊น์ง ์ฒ๋ฆฌ๊ฐ๋ฅํ๋ฐ batch mode๋ก ๋๋ฆฌ๋ฉด ์ต๋ 3000๊ฐ๊น์ง ์ฒ๋ฆฌ๊ฐ๋ฅํ๋ค.
operationํ๋๋น ์ต๋ 300๊ฐ๊น์ง ์ฒ๋ฆฌ๊ฐ๋ฅํ๋ฏ๋ก 4 messages per operation์ด๋ฉด ์ต๋ 1200๊ฐ ์ฒ๋ฆฌ๊ฐ๋ฅํ๋ค.
#7. Placement Group
Cluster Placement Group : ๋ชจ์ฌ ์์ด์ ๋คํธ์ํฌ์ ์ผ๋ก ์ด์ ์กด์ฌ
Partition Placement Group : partition๊ฐ์ ๋ ๋ฆฝ์ ์ด๋ partition๋ด๋ถ์ ์ผ๋ก๋ ๋คํธ์ํฌ์ ์ผ๋ก ์ด์ ์กด์ฌ
Spread Placement Group : high availability
#8. own custom DNS service
own custom DNS service๋ฅผ ์ฌ์ฉํ๋ฉด Route 53์ ์ฌ์ฉํ ํ์๊ฐ ์๋ค.
#9. UDP protocol, fast regional failover if an AWS Region goes down
Global Accelerator improves performance for a wide range of applications over TCP or UDP by proxying packets at the edge to applications running in one or more AWS Regions.
Global Accelerator is a good fit for non-HTTP use cases, such as gaming (UDP), IoT (MQTT), or Voice over IP, as well as for HTTP use cases that specifically require static IP addresses or deterministic, fast regional failover.
#10. ASG
ํจ์น ์ ๋ต
ReplaceUnhealthy๋ฅผ suspendํด์ ํจ์น์๋ฃ ๋ ๋๊น์ง ์ธ์คํด์ค๋ฅผ replaceํ์ง ์๋๋ก ํจ
standby state๋ฅผ ์ฌ์ฉํด์ ํจ์น์๋ฃํจ. standby state์ ์ธ์คํด์ค๋ asg์์ ํฌํจ๋๋ ํธ๋ํฝ์ ๋ฐ์ง ์์
#11. ASG policy
scaling policy
cpu 50%๋ฅผ ํ๊ฒ์ผ๋ก ์ธ์คํด์ค์ ๊ฐ์๋ฅผ ์ ์งํ๊ณ ์ถ๋ค๋ฉด target tracking policy
์ต์ ์ , ์ต๊ณ ์ ๊ธฐ์ค์ผ๋ก ์ธ์คํด์ค์ ๊ฐ์๋ฅผ ์ ์งํ๊ณ ์ถ๋ค๋ฉด step, simple policy
step๊ณผ simple policy์ ์ฐจ์ด์
simple์ ์ธ์คํด์ค๊ฐ ์์ฑ๋๊ฑฐ๋ ์ ๊ฑฐ๋๋ ๊ณผ์ ์ค์์ policy๋ฅผ ์คํํด ์ถ๊ฐ์ ์ธ ์์ ์ ๋ฐ๋์ํค์ง ์๋๋ค.
step์ ์ธ์คํด์ค๊ฐ ์์ฑ๋๊ฑฐ๋ ์ ๊ฑฐ๋๋ ๊ณผ์ ์ค์์๋ policy๋ฅผ ์คํํด ์ถ๊ฐ์ ์ธ ์์ ์ ๋ฐ๋์ํจ๋ค.
#12. ์ ์ฉํ Architecture
Kinesis Firehose๋ DynamoDB์ ๋ฐ์ดํฐ๋ฅผ ์ ๋ฌํ ์ ์๋ค.
SQS Standard > Lambda Batch > DynamoDB
Sudden Traffic Spike ๊ฐ๋น๊ฐ๋ฅํ ์๋น์ค(์ค์ ํ ์คํธ1 - 36)
API GATEWAY(Token Buffer๋ก request์ ๊ฐ์ ์ ํ ๊ฐ๋ฅ) > SQS(Buffer ์ญํ ๊ฐ๋ฅ) > KINESIS(Buffer ์ญํ ๊ฐ๋ฅ)
Major traffic spikes ๊ฐ๋น ๊ฐ๋ฅ, store the processed updates in a highly available database, minimize the management overhead(์๋ฒ๋ฆฌ์ค์ ๊ฐ๊น์ด)
Kinesis Data Streams(order ๋ณด์ฅ, ๋ง์ ๋ฐ์ดํฐ ์์ค๋ก๋ถํฐ ๋ฐ์ดํฐ ์์ง๊ฐ๋ฅ) > Lambda function > DynamoDB
#13. HDD๋ Boot Volume์ด ๋ ์ ์๋ค.
#14. Aurora Read Replica failover priority
tier๊ฐ ๋ฎ์ ๊ฒ ์ฐ์ (ex. 1์ด 15๋ณด๋ค ์ฐ์ ์์๊ฐ ๋์)
tier๊ฐ ๊ฐ๋ค๋ฉด ์ฉ๋์ด ๋์ ๊ฒ ์ฐ์
tier์ ์ฉ๋์ด ๊ฐ๋ค๋ฉด arbitrary๋ก ๋ฐฐ์
#15. Instance Store : fleet of instances์์ instanceํ๋๊ฐ ๋ค์ด๋์ด๋ ๋ค๋ฅธ ์ธ์คํด์ค๊ฐ ๋์ฒด๊ฐ๋ฅํ๋ค๋ ์ ์์ resilient architecture์ด๋ค.
fleet of instances๊ฐ ์์ ๋ instance store๋ฅผ ์ฌ์ฉํ๋ฉด instance๊ฐ์ instance store์ ๋ฐ์ดํฐ๋ฅผ ๊ณต์ ๊ฐ๋ฅ
๋ฐ๋ผ์ ์ธ์คํด์ค ํ๋๊ฐ ๋ค์ด๋์ด๋ ๋ค๋ฅธ ์ธ์คํด์ค๊ฐ ๋ค์ด๋ ์ธ์คํด์ค๊ฐ ์ฌ์ฉํ๋ ๋ฐ์ดํฐ์ ์ ๊ทผ ๊ฐ๋ฅ
์์ ์คํ ๋ฆฌ์ง
ํธ์คํธ ์ปดํจํฐ์ ๋ฌผ๋ฆฌ์ ์ผ๋ก ๋ถ์ฐฉ๋จ
load-balanced pool of web servers : best practice
Instance store volumes are included as part of the instance's usage cost : instance์ฒ๋ผ ์ทจ๊ธ๋์ด instance ๋น์ฉ์ ์ฒญ๊ตฌ๋จ
#16. ALB๋ ์ฌ์ค IP, Instance, lambda๋ฅผ ๋์์ผ๋ก ํ๊ณ ๊ณต์ธ IP๋ก ๋ผ์ฐํ ํ์ง ์๋๋ค.
#17. Direct Connect
์ค์ ํ ์คํธ1 - 27๋ฒ, ๋ค์ ๋ณด๋ ๊ฒ ์ถ์ฒ
Site-to-site VPN cannot provide low latency and high throughput connection : Internet-based connectivity๋ฅผ ์ฌ์ฉํ๋ VPN์ ์๋๊ฐ ๋น ๋ฅด์ง ์๋ค.
Site-to-site VPN : ํ์ง๋ง Immediate need์ ๋์๊ฐ๋ฅํ๋ค.
Site-to-site VPN : IPSec to establish encrypted network connectivity between your intranet and Amazon VPC over the Internet > Internet-based connectivity์ธ ๊ฒ์ ๋ณด์ฌ์ค. encrypted network์ด์๋ง public network์ธ internet์ ์ฌ์ฉํ๋ค.
AWS Direct Connect by itself cannot provide an encrypted connection between a data center and AWS Cloud : Direct Connect ์์ฒด๋ encrypted connection๊ณผ ๊ด๋ จ์ด ์๋ค. encrypted network์ ๊ด๋ จ์ด ์์ง๋ง, private network๋ฅผ ์ฌ์ฉํ๋ค.
Site-to-site VPN + Direct Connect : IPsec-encrypted(by VPN) private connection(by DC) that also reduces network costs(์ธํฐ๋ท์์ ๋ผ์ฐํ ํ๋ ๊ฒ๋ณด๋ค private network๋ฅผ ์ฌ์ฉํด ๋ผ์ฐํ ํ๋ฉด ๋ ๋น ๋ฅด๊ธฐ ๋๋ฌธ์), increases bandwidth throughput(DC๋ฅผ ์ฌ์ฉํ๋ฉด Increase bandwidth throughput์ด ๊ฐ๋ฅํจ)
#18. Transit Gateway
๋ชจ๋ aws ์๋น์ค ์ค์์ ์ ์ผํ๊ฒ multicast์ง์
#19. Kinesis Data Steams
SNS+SQS Fan Out๋ณด๋ค multiple applications to consume the same streamํ๋๋ฐ์ ์ ๋ฆฌํจ
๋ํ์ Use case 3๊ฐ์ง
Routing related records to the same record processor : ์ฐ๊ด๋ ๋ ์ฝ๋๋ฅผ ๊ฐ์ ํ๋ก์ธ์์ ์ ๋ฌ
Ordering : immediate์ด๋ a few hours later์ด๋ ๋ฐ์ดํฐ์ ์์๋ณด์ฅ๊ฐ๋ฅ
Ability for multiple applications to consume the same stream concurrently. For example, you have one application that updates a real-time dashboard and another that archives data to Amazon Redshift. You want both applications to consume data from the same stream concurrently and independently. : Kinesis Data Steams์ multiple applications to consume the same stream in real-time์ ์ต์ ํ๋จ
partition key๋ถ๋ถ ๊ฐ์์๋ฃ ๋ค์ ๋ณผ ๊ฒ
#20. Route53 : Host-based Routing
You can route a client request based on the Host field of the HTTP header allowing you to route to multiple domains from the same load balancer. : ์๋ก ๋ค๋ฅธ ๋๋ฉ์ธ ๊ฐ ๋ผ์ฐํ ๋ ๊ฐ๋ฅํ๋ค.
#21. FSx for Windows File Server(์ค์ ํ ์คํธ1 - 41)
support DFS(Distributed File System), SMB Protocol ์ฌ์ฉ
S3 objects as files and does not allow you to write changed data back to S3.
user quotas, end-user file restore
AWS Managed Microsoft AD, FSx for Lustre๋ DFS๋ฅผ supportํ์ง ์๋๋ค.
Amazon FSx๋ ์จํ๋ ๋ฏธ์ค Microsoft Active Directory ๋ฐ AWS Microsoft Managed AD์ ํตํฉ๋๋ค.
AWS DataSync๋ฅผ ํตํ ๊ฐ๋จํ๊ณ ์ํํ ๋ง์ด๊ทธ๋ ์ด์ : AWS DataSync๋ฅผ ์ฌ์ฉํ๋ฉด ์จ ํ๋ ๋ฏธ์ค ํ์ผ ์์คํ ์ Amazon FSx์ ์์ ๊ด๋ฆฌํ Windows ์คํ ๋ฆฌ์ง๋ก ์ฝ๊ฒ ์ด๋ํ ์ ์๋ค. AWS DataSync์์ ํตํฉ์ ์ธํฐ๋ท ๋๋ AWS Direct Connect๋ฅผ ์ฌ์ฉํ๋ฉด ๋ฐ์ดํฐ ๋ณต์ฌ๋ฅผ ์๋ํ ๋ฐ ๊ฐ์ํํ ์ ์๋ค. ๋ํ It is natively integrated with Amazon S3, Amazon EFS, Amazon FSx for Windows File Server, Amazon CloudWatch, and AWS CloudTrail.
#22. AMI๋ ebs snapshot์ด ๋ฒ ์ด์ค์ด๋ค.
๋ฐ๋ผ์ ๋ฆฌ์ ๊ฐ ami copy์ ๋ณต์ฌํ ๋ฆฌ์ ์๋ AMI, EBS Snapshot 2๊ฐ๊ฐ ์์ฑ๋๋ค.
#23 s3๋ prefix๋น 3,500 PUT/COPY/POST/DELETE or 5,500 GET/HEAD์ด ๊ฐ๋ฅํ๊ณ prefix์ ๊ฐ์๋ ๋ฌด์ ํ์ด๋ค.
๋ฐ๋ผ์ customer-specific custom prefixes์ ์ฌ์ฉํ๋ฉด ์ฌ์ค์ request์ ์ ํ์ด ์๋ค.
#24 Kinesis
Kinesis Data Streams๋ firehose์ฒ๋ผ intermediary Lambda function์ ์ฌ์ฉํ ์ ์๋ค.
Kinesis Data Analytics๋ streams์ ๋ฌ๋ฆฌ ๋ค์ํ ์์ค๋ก๋ถํฐ ๋ฐ์ดํฐ๋ฅผ ์ ๋ฌ ๋ฐ์ ์ ์๊ณ , ๋ณดํต data streams ๋๋ data firehose๋ก๋ถํฐ ์ ๋ฌ๋ฐ๋๋ค.
#25 S3์์ standard๋ฅผ ์ ์ธํ๋ฉด minimum storage duration์ด ์ต์ 30์ผ์ผ๋ก ๊ทธ ์ดํ์ data transition์ด ๊ฐ๋ฅํ๋ค.
#26 Lambda๋ default๋ก ๋์์ ์ต๋ 1000๊ฐ๊น์ง ์คํํ๊ณ ๊ทธ ์ด์์ ์ํ๋ค๋ฉด aws support์ ๋ฌธ์ํด์ผ ํ๋ค.
#27 WAF - Geo match conditions
configure a whitelist that allows only viewers in those countries.
configure a blacklist so that end-users from those countries are blocked from downloading their software.
WAF๋ฅผ ์ฌ์ฉํ๋ฉด cloudfront๊ฐ edge๋ ๋ฒจ(doesn't belong to VPC)์์ geo restrictionํ๋ ๊ฒ๊ณผ ๋ฌ๋ฆฌ VPC - ALB๋ ๋ฒจ์์ geo match ํ์ดํธ๋ฆฌ์คํธ, ๋ธ๋๋ฆฌ์คํธ๋ฅผ ์ฌ์ฉํ ์ ์๋ค.
#28 FSx For Lustre(์ค์ ํ ์คํธ1 - 52)
HPC, fast storage : FSx For Lustre๋ ๊ณ ์ฑ๋ฅ ํ์ผ ์์คํ ์ด๋ค.
FSx for Lustre provides the ability to both process the 'hot data' in a parallel and distributed fashion as well as easily store the 'cold data' on Amazon S3
FSx for Lustre integrates with Amazon S3
๋ค๋ฅธ ์๋น์ค(ex. EFS์ ๋นํด ๋น์ธ๋ค)
์ปดํจํ ํ์๋งํผ ๋น ๋ฅธ ์คํ ๋ฆฌ์ง ์ฑ๋ฅ์ ์ํด์ ์ฌ์ฉ
#29 ์ธ์คํด์ค ๋ ๋ฒจ์ ์ก์ธ์ค ์ ํ
VPC security groups
IAM policy
EFS Access Points
When an NFS client mounts an EFS file system without using an access point, the user ID and group ID provided by the client is trusted.
You can use EFS access points to override user ID and group IDs used by the NFS client.
When users attempt to access files and directories, Amazon EFS checks their user IDs and group IDs to verify that each user has permission to access the objects
์ฆ access point๋ฅผ ์ฌ์ฉํ์ง ์์ผ๋ฉด ๊ถํ ์ฒดํฌ๋ฅผ ํ์ง ์๋๋ฐ access point๋ฅผ ์ฌ์ฉํ๋ฉด ๊ถํ ์ฒดํฌ๋ฅผ ํ๋ค.
#30 ECS Cost
ECS with EC2 launch type is charged based on EC2 instances and EBS volumes used.
ECS with Fargate launch type is charged based on vCPU and memory resources that the containerized application requests
#31 EFS๋ก์ region๊ฐ ์ ๊ทผ
The spreadsheet on the EFS file system can be accessed in other AWS regions by using an inter-region VPC peering connection
#32 ec2 user data
ec2 user data๋ ์ต์ด ๋ถํ ์์๋ง ์คํ๋จ
ec2 user data๋ default๋ก root ๊ถํ์ ๊ฐ์ง
#33 By default, an S3 object is owned by the AWS account that uploaded it. So the S3 bucket owner will not implicitly have access to the objects written by Redshift cluster : a๊ณ์ ์ด b๊ณ์ ์ s3๋ฒ์ผ์ ํ์ผ์ ์ ๋กฃ๋ํ๋ฉด ๋ฒ์ผ์ ์์ ํ b๊ณ์ ์ ๊ธฐ๋ณธ์ ์ผ๋ก ๊ทธ ํ์ผ์ ์ ๊ทผํ ์ ์๋ค. ํ์ผ์ ์ฌ๋ฆฐ ๊ณ์ ์ด ์๋๊ธฐ ๋๋ฌธ์ด๋ค.
#34 IAM permission boundary. They can only be applied to roles or users, not IAM groups
#35 ASG๊ฐ unhealthy instance๋ฅผ terminateํ์ง ์์ ๋
The health check grace period for the instance has not expired : grace period๋ผ๊ณ terminateํ์ง ์๋ ์ ์ ๊ธฐ๊ฐ์ด ์๋ค.
The instance maybe in Impaired status - ์์๋ status์ด๋ฉด recoverํ ์๊ฐ์ ์ฃผ๊ธฐ ๋๋ฌธ์ terminateํ์ง ์๋๋ค.
The instance has failed the ELB health check status - By default, Amazon EC2 Auto Scaling doesn't use the results of ELB health checks to determine an instance's health status when the group's health check configuration is set to EC2 : ์ฆ asg๋ ๊ธฐ๋ณธ์ ์ผ๋ก elb health check๋ณด๋ค ec2 health check์ ์ฐ์ ์์๋ก ๋๊ธฐ ๋๋ฌธ์ elb health check์ด failํ๋๋ผ๋ ec2 health๋ ์ ์์ผ ์ ์๋ค๋ ๊ฒ์ด๋ค.
#36 cognito user pool vs cognito identity pool
๋ ๋ค ์์ ๋ก๊ทธ์ธ, SAML์ ์ง์ํ๋ค.
Cognito user pool
์ ํ๋ฆฌ์ผ์ด์ ์ ํ์๊ฐ์ , ๋ก๊ทธ์ธ, ์์ ๋ก๊ทธ์ธ ๊ธฐ๋ฅ์ ๋ถ์ฐฉ๊ฐ๋ฅ. ์ฆ ์ฑ์ ์ฌ์ฉํ ๋ ์ธ์ฆ์ ์ํด์ ์ฌ์ฉํ๋ค.
Cognito identity pool
Amazon S3 ๋ฐ DynamoDB๊ฐ์ aws ์๋น์ค์ ์ก์ธ์คํ๊ธฐ ์ํด ์์ ์๊ฒฉ ์ฆ๋ช ์ ์ป์ ๋ ์ฌ์ฉํ๋ค.
#37 Use Cognito Authentication via Cognito User Pools for your Application Load Balancer : true
Use Cognito Authentication via Cognito User Pools for your CloudFront distribution : You cannot directly integrate Cognito User Pools with CloudFront distribution as you have to create a separate Lambda@Edge function to accomplish the authentication via Cognito User Pools.
#38 Kinesis
Kinesis firehose๋ fully managed์ธ๋ฐ ๋นํด Kinesis Data Streams๋ shard๋ฅผ provisionํด์ผ ํ๋ค.
#39 Spot Fleet Request๋ spot instance๋ค์ ์์ฒญํ๋ ๊ฒ์ด์ง asg์ฒ๋ผ ์ ๋์ ์ผ๋ก ์ธ์คํด์ค๋ฅผ terminateํ๊ณ createํ๋ ๊ฒ์ด ๋ถ๊ฐ๋ฅํ๋ค.
#40 video๋ rds๊ฐ ์๋ s3์ ์ ํฉ
#41 MAX I/O performance mode in EFS
์ง์ฐ ์๊ฐ์ด ๋์ด๋์ง๋ง, ๋ณ๋ ฌํ๋ ์ ํ๋ฆฌ์ผ์ด์ ์ ํนํ๋๊ณ , ๋ ๋ง์ throughput ์ง์
#42 Kinesis firehose์ source๋ฅผ kinesis data streams๋ก ์ฌ์ฉ์ค์ด๋ผ๋ฉด, kinesis agent๋ kinesis firehose์ directํ๊ฒ ๋ฐ์ดํฐ๋ฅผ ์ ๋ฌํ ์ ์๋ค.
๋ฐ๋ผ์ ์ด ๊ฒฝ์ฐ kinesis agent๋ ๋ฐ์ดํฐ๋ฅผ kinesis data streams์ ์ถ๊ฐํด์ผ ํ๋ค.
๊ธฐ๋ณธ์ ์ผ๋ก kinesis agent๋ kinesis firehose, kinesis data streams์ ๋ฐ์ดํฐ๋ฅผ ์ ๋ฌํ ์ ์๋ค.
#43 can't move data directly from Snowball into a Glacier Vault or a Glacier Deep Archive Vault. You need to go through S3 first
์ค๋ ธ์ฐ๋ณผ์์ glacier๋ก ๋ฐ์ดํฐ๋ฅผ ์ฎ๊ธฐ๋ ค๋ฉด ๋จผ์ s3๋ก ์ฎ๊ธด ํ lifecycle๋ก ์ฎ๊ธฐ๋ ๊ฒ์ด ์ผ๋ฐ์ ์ด๋ค.
#44 You can only use a launch template to provision capacity across multiple instance types using both On-Demand Instances and Spot Instances to achieve the desired scale, performance, and cost
launch template : versioning๊ฐ๋ฅ, provision capacity across multiple instance types using both On-Demand Instances and Spot Instances๊ฐ๋ฅ
launch configuration : ์ 2๊ฐ์ง ๋ถ๊ฐ๋ฅ
#45 Create an encrypted snapshot of the database, share the snapshot, and allow access to the AWS Key Management Service (AWS KMS) encryption key
You can share the AWS Key Management Service (AWS KMS) customer master key (CMK) that was used to encrypt the snapshot with any accounts, that you want to be able to access the snapshot. You can share AWS KMS CMKs with another AWS account by adding the other account to the AWS KMS key policy.
CMK์ ๋ํ ์ก์ธ์ค๋ ์ฌ๋ฌ ๊ณ์ ๊ฐ์ ๊ณต์ ํ๋ ๊ฒ๋ ๊ฐ๋ฅํ๋ค.
Making an encrypted snapshot of the database by CMK will give the auditor a copy of the database ์๋ํ๋ฉด AWS KMS key policy๋ฅผ ๋ฐ๊ฟ์ ํค์ ๋ํ ์ ๊ทผ ๊ถํ์ ๊ณต์ ํ๊ธฐ ๋๋ฌธ์ด๋ค.
#46 "aws:RequestedRegion": "eu-west-1"์ api call์ด ๋ง๋ค์ด์ง ๊ณณ ๊ธฐ์ค์ด ์๋๋ผ, instance๊ฐ ์ด๋ ๋ฆฌ์ ์ ์กด์ฌํ๋์ง๋ฅผ ๊ธฐ์ค์ผ๋ก ํ๋ค.
์๋๋ policy์ด๋ค.
#47 ๋ธ๋ฃจ ๊ทธ๋ฆฐ ๋ฐฐํฌ๋ฅผ ํ๋๋ฐ DNS์บ์ฑ์ด ๋ฐ์ํ ๊ฒฝ์ฐ, Global Accelerator๊ฐ dns ์บ์ฑ์ ํด๊ฒฐํด ์ค ์ ์๋ค.
Use AWS Global Accelerator(multi-Region solution) to distribute a portion of traffic to a particular deployment
"AWS Global Accelerator๋ฅผ ์ฌ์ฉํ๋ฉด ํด๋ผ์ด์ธํธ ๋๋ฐ์ด์ค์ ์ธํฐ๋ท ๋ฆฌ์กธ๋ฒ์์ DNS ์บ์ฑ์ ์ข ์๋์ง ์๊ณ ํธ๋ํฝ์ ์ ์ง์ ์ผ๋ก ๋๋ ๋ชจ๋ ํ ๋ฒ์ ์ด๋ํ ์ ์์ผ๋ฉฐ, ํธ๋ํฝ ๋ค์ด์ผ ๋ฐ ๋์ ๊ฐ์ค์น ๋ณ๊ฒฝ์ ๋ช ์ด ๋ด์ ์ ์ฉ๋ฉ๋๋ค."
#48 Secrets manager๋ ๋ณ์ ์๋ ๋กํ ์ด์ ๊ธฐ๋ฅ์ ์ง์ํ์ง๋ง, SSM Paramter Store๋ ์๋ ๋กํ ์ด์ ๊ธฐ๋ฅ์ ์ง์ํ์ง ์๊ณ , ์๋์ผ๋ก ๋๋ ค์ผ ํ๋ค.
#49
A developer needs to implement a Lambda function in AWS account A that accesses an Amazon S3 bucket in AWS account B.
์ ์ํฉ์ ํ์ํ ๋ ๊ฐ์ง ์ค์ ์ ์๋์ ๊ฐ๋ค.
S3 ๋ฒํท์ ์ก์ธ์คํ ์ ์๋ ๋๋ค ๊ธฐ๋ฅ์ ๋ํ IAM ์ญํ ์ ๋ง๋ ๋ค.
IAM ์ญํ ์ ๋๋ค ๊ธฐ๋ฅ์ ์คํ ์ญํ ๋ก ์ค์ ํฉ๋๋ค.
๋ฒํท ์ ์ฑ ์ด ๋๋ค ํจ์์ ์คํ ์ญํ ์ ๋ํ ์ก์ธ์ค ๊ถํ๋ ๋ถ์ฌํด์ผ ํ๋ค.
#50 Instance๊ฐ ์ข ๋ฃ๋ ํ์๋ EBS Volume์ ์ ์งํ๋ ๋ฐฉ๋ฒ
Set the DeleteOnTermination attribute to false
์์ ๋ฌ๋ฆฌ ec2 hibernate๋ in-memory state๋ฅผ ์ ์งํ๊ฒ ํ๋ค. : hibernate๋ฅผ ์ฌ์ฉํ๋ฉด in-memory์ ๋ด์ฉ์ ebs์ ์ ์ฅํ๊ธฐ ๋๋ฌธ์ ๊ฐ๋ฅํ ๊ฒ์ด๋ค.
#51 SQS์์ group id๋ฅผ ์ฌ์ฉํ์ง ์์ผ๋ฉด consumer๋ only one์ด๋ค.
#52 Kinesis Data Streams์ ๋นํด SQS FIFO๋ consumers๋ฅผ ๋๋ฆฌ๊ธฐ ํจ์จ์ ์ด๋ค(ํ SQS์ ์ต๋ 100๊ฐ์ consumer).
#53 IAM Account level, User level ์ก์ธ์ค ๊ถํ
https://docs.aws.amazon.com/ko_kr/IAM/latest/UserGuide/tutorial_cross-account-with-roles.html(IAM ์ญํ ์ ์ฌ์ฉํ AWS ๊ณ์ ๊ฐ ์ก์ธ์ค ๊ถํ ์์) ์ฐธ๊ณ
์ ๋ฌธ์๋ฅผ ์ฝ์ด๋ณด๋ฉด ๊ณ์ ๊ฐ ๊ถํ ์์์ IAM Policy๋ฅผ ์ฌ์ฉํ๋ค.(AssumeRole, Security Token Service)
#54 S3 Account level, User level ์ก์ธ์ค ๊ถํ
IAM Policies
No
Yes
ACLs
Yes
No
Bucket Policies
Yes
Yes
Bucket Polices๋ ip๋ฅผ ๊ธฐ์ค์ผ๋ก๋ ์ ํ ๊ฐ๋ฅ
#55 Elastic Load Balancing does not work across regions : ELB๋ ํ ๋ฆฌ์ ์ ๊ตญํ๋ ์๋น์ค์ด๋ค.
#56
Does S3 bucket policy override IAM policy? : S3 bucket policy๊ฐ iam policy๋ฅผ ๋ฌด์ํ๊ณ ๋ฎ์ด์์ฐ๋ ๊ฒ์ด ๊ฐ๋ฅํ๊ฐ?
Yes it can indeed override the policy, but only where it uses a Deny. If it includes an Allow but the IAM policy includes a Deny this will not evaluate as Allow. : deny์ ํํด์ ๊ฐ๋ฅํ๋ค. iam policy๊ฐ deny์ธ๋ฐ bucket policy๊ฐ allowํ๋ค๊ณ ํด์ override๋์ง ์๋๋ค.
#57 storage gateway๋ on-premise๋ฐ cloud์ ํ์ด๋ธ๋ฆฌ๋ ํ๊ฒฝ์ ์ฌ์ฉํ๋ ์ฌ๋ด์ ๋ฐ์ดํฐ๋ฅผ ์บ์ฑํ๋ ๊ธฐ๋ฅ๋ ํ๋ค.
#58 PostgreSQL์ ๊ธฐ๋ณธ ํฌํธ๋ 5432์ด๋ค.
#59 Batch job์ spot instance๊ฐ ๋น์ฉ๋ฉด์์ ์ต์ ํ๋์ด ์๋ค.
#60 Partition placement group์ Hadoop, ์นด์ฐ๋๋ผ, ์นดํ์นด ๊ฐ์ ๋๊ท๋ชจ ๋ฐ์ดํฐ ๋ถ์ฐ ์์ ์ ์ฌ์ฉ๋๋ค.
#61 ASG Default Termination Policy
Find the AZ which has the most number of instances : ๋จผ์ ๊ฐ์ฅ ๋ง์ ์ธ์คํด์ค๊ฐ ์๋ az๋ฅผ ์ฐพ๋๋ค.
1์์๋ก ์ ์ผ ๋จผ์ terminate๋๋ ๋์ : ์ค๋๋ launch configuration
2์์๋ก terminate๋๋ ๋์ : ์ค๋๋ launch template
3์์๋ก terminate๋๋ ๋์ : closest to next billing hour - ์ด๋ ์๊ฐ ๋จ์๋ก ์ฒญ๊ตฌ๋๋ linux, ubuntu ec2 usage cost๋ฅผ ์ค์ฌ์ค๋ค.
#62 CloudFormation์ ๋ฆฌ์์ค๋ฅผ ํ๋ก๋น์ ๋ํ๋ ๋ฐ ์๊ฐ์ด ๊ฑธ๋ฆฌ๊ธฐ ๋๋ฌธ์ ํน์ ์ฌ์ฉ ์ฌ๋ก์ ์ต์๋์ ๋ค์ดํ์์ด ํ์ํ ๊ฒฝ์ฐ์๋ ์ ์ ํ ์๋ฃจ์ ์ด ์๋๋ค.
#63 Reserved Instance vs Spot Instance
Reserved Instance๋ ์ง์์ ์ธ ์ฌ์ฉ์ ํจ์จ์
Spot Instance๋ monthly work์ ํจ์จ์ ์ด๋ค.
๊ทธ๋ฌ๋ monthly work๋ผ๋ ์์ ์ ์ค๋จํ๋ฉด ์๋๋ ๊ฒฝ์ฐ ํน์ ํน์ ์๊ฐ ๋ด์ ์๋ฃํด์ผ ํ๋ ๊ฒฝ์ฐ๋ ์ ํฉํ์ง ์๋ค.
Spot Instances are well-suited for data analysis, batch jobs, background processing, and optional tasks
Amazon EC2 needs the capacity back์ผ ๋, Spot instance๊ฐ ์ข ๋ฃ๋ ์ ์๋๋ฐ ์ด ๋ ๊ณต์๋ฌธ์์์๋, "Amazon EC2 automatically resubmits a persistent Spot Instance request after the Spot Instance associated with the request is terminated"๋ผ๊ณ ๋งํ๋ฉฐ ๋ค๋ฅธ spot instance๋ฅผ ์๋์ผ๋ก ์์ฒญํ๋ค.
์ฆ Amazon EC2 needs the capacity back ๋๋ Spot price exceeds the maximum price for your request์ผ ๋ spot instance๋ terminate๋๋ค.
#64 Dedicated Host๋ Dedicated Instance์ ๋นํด cost๊ฐ ๋ง์ด ๋ค์ด cost-effectiveํ์ง ์๋ค.
#65 ์ค์ ํ ์คํธ 2 - 63๋ฒ bucket policy ๋ฌธ์ ์์
#66 Shared Service VPC
์ค์ ํ ์คํธ 2 - 65๋ฒ
ํ ํ์ฌ๊ฐ AWS ๊ณ์ ์ ์ฌ๋ฌ ๊ฐ ์ด์ํ๊ณ ์์ผ๋ฉฐ AWS Transit Gateway๋ฅผ ์ฌ์ฉํ์ฌ ํ๋ธ ์ค ์คํฌํฌ ๋ฐฉ์์ผ๋ก ์ด๋ค ๊ณ์ ์ ์ํธ ์ฐ๊ฒฐํ์ต๋๋ค. ๋คํธ์ํฌ ๋ถ๋ฆฌ๋ฅผ ์ฉ์ดํ๊ฒ ํ๊ธฐ ์ํด ์ด๋ฌํ AWS ๊ณ์ ์ ์ฒด์ VPC๊ฐ ํ๋ก๋น์ ๋๋์์ต๋๋ค. VPC์ ์ํฌ๋ก๋์ ํ์ํ ์๋น์ค์ ๋ํ ๊ณต์ ์ก์ธ์ค๋ฅผ ์ ๊ณตํ๋ฉด์ ๊ด๋ฆฌ ์ค๋ฒํค๋์ ๋น์ฉ์ ๋ชจ๋ ์ค์ผ ์ ์๋ ์๋ฃจ์ ์ ๋ฌด์์ ๋๊น?
Sharing resources from a central location instead of building them in each VPC may reduce administrative overhead and cost : ๊ฐ vpc์์ ์์์ ๋ง๋ค์ด ๊ณต์ ํ๋ ๊ฒ์ด ์๋๋ผ Shared Service VPC๋ฅผ ๋ง๋ค์ด ์ฌ์ฉํ๋ค.
https://aws.amazon.com/ko/blogs/architecture/reduce-cost-and-increase-security-with-amazon-vpc-endpoints/(Amazon VPC ์๋ํฌ์ธํธ๋ก ๋น์ฉ ์ ๊ฐ ๋ฐ ๋ณด์ ๊ฐํ)
#67 When you publish a high-resolution metric, CloudWatch stores it with a resolution of 1 second, and you can read and retrieve it with a period of 1 second, 5 seconds, 10 seconds, 30 seconds, or any multiple of 60 seconds : high-resolution metric์ ์ฌ์ฉํ๋ฉด 1์ด ๊ฐ๊ฒฉ์ผ๋ก ์๊ณ , 1~60์ด ๊ฐ๊ฒฉ์ผ๋ก retreiveํ ์ ์๋ค.
#68 EC2 Detailed Monitoring์ 1๋ถ ๊ฐ๊ฒฉ์ผ๋ก metric ๋ชจ๋ํฐ๋ง ๊ฐ๋ฅ
#69 Endpoint on Route 53 Resolver
Create an inbound endpoint on Route 53 Resolver and then DNS resolvers on the on-premises network can forward DNS queries to Route 53 Resolver via this endpoint : Route 53 Resolver์ ์ธ๋ฐ์ด๋ ์๋ํฌ์ธํธ๋ ์จํ๋ ๋ฏธ์ค์ DNS resolver๊ฐ Route 53 Resolver์ ์ฟผ๋ฆฌ๋ฅผ ์์ฒญํ ์ ์๊ฒ ํ๋ค.
Create an outbound endpoint on Route 53 Resolver and then Route 53 Resolver can conditionally forward queries to resolvers on the on-premises network via this endpoint : Route 53 Resolver์ ์์๋ฐ์ด๋ ์๋ํฌ์ธํธ๋ ์กฐ๊ฑด์ ์ผ๋ก ์จํ๋ ๋ฏธ์ค ๋คํธ์ํฌ์ DNS resolver์ ์ฟผ๋ฆฌ๋ฅผ ์์ฒญํ ์ ์๊ฒ ํ๋ค.
inbound endpoint ๋ฐฉํฅ : on-premises DNS resolvers > Route 53 Resolver
outbound endpoint ๋ฐฉํฅ : Route 53 Resolver > on-premises DNS resolvers
์ฐธ๊ณ ๋ก, dns resolver๊ฐ dns server๋ก ์ฟผ๋ฆฌ๋ฅผ ๋ณด๋ด ์๋ต์ ์์ฒญํ๋๋ฐ dns resolver > local dns server > (root dns server, tld dns server, sld dns server) ์ด ์์๋ก ์์ฒญ์ด ์ด๋ฃจ์ด์ง๋ค๊ณ ๋ณด๋ฉด ๋๋ค.
#70 Aurora Global Database
Short Recovery Time(RTO)์ ํนํ
Managed planned failover โ ์๋์ผ๋ก failover๋ฅผ ์คํ
Unplanned failover - ์ง์ failover์คํํ๊ธฐ ๋๋ฌธ์ RTO๊ฐ ๊ธธ์ด์ง ์ ์์
#71 AWS Elastic Beanstalk
full control over the AWS resources powering your application and can access the underlying resources at any time
#72 Dedicated Hosts enable you to use your existing server-bound software licenses like Windows Server and address corporate compliance and regulatory requirements : dedicated host๋ dedicated instance์ ๋ฌ๋ฆฌ ์จํ๋ ๋ฏธ์ค ์๋ฒ์ ์ํํธ์จ์ด ๋ผ์ด์ผ์ค๋ฅผ dedicated host์ ๋๊ฐ์ด ์ ์ฉ์ํฌ ์ ์๋ค.
#73 Configure an Amazon CloudWatch alarm that triggers the recovery of the EC2 instance, in case the instance fails. The instance can be only configured with EBS volume - The recover action is supported only on instances that have EBS volumes configured on them, instance store volumes are not supported for automatic recovery by CloudWatch alarms. : ebs๊ฐ ์๋ instance store๋ฅผ ์ฌ์ฉํ๋ ๊ฒฝ์ฐ ์๋ ๋ณต๊ตฌ๋ฅผ ์ง์ํ์ง ์๋๋ค.
#74 ALB์ EC2 Instances๋ฅผ ์ฌ์ฉํ๋๋ฐ ๋๋ฌด ๋ง์ ALB์ฌ์ฉ์ผ๋ก ๊ตฌ์กฐ๊ฐ ๋ณต์กํด์ก์ ๊ฒฝ์ฐ
The architecture has now become complex with too many ALBs in multiple AWS Regions. Security updates, firewall configurations, and traffic routing logic have become complex with too many IP addresses and configurations.
ํด๊ฒฐ์ฑ : Launch AWS Global Accelerator and create endpoints for all the Regions. Register the ALBs of each Region to the corresponding endpoints
#75 ALB, ASG ์๋ elastic ip๋ฅผ ํ ๋นํ ์ ์๋ค.
#76 ์ค์ ํ ์คํธ 3 - 11๋ฒ
ASG๋ก SQS์ ํ๋ฅผ ๋ฐ์ ์ฌ์ฉํ๋ ์ํคํ ์ฒ๊ฐ ์๋๋ฐ, a sudden spike in orders received๋ฅผ ์ด๋ป๊ฒ ๊ฐ๋นํ ๊ฒ์ธ๊ฐ
Use a target tracking scaling policy based on a custom Amazon SQS queue metric.
ํ์ง๋ง NumberOfMessages๋ฅผ SQS metric์ผ๋ก ์ค์ ํ๋ค๋ฉด, sqs์ ๋ฉ์์ง์ ์๊ฐ ๋ณ๊ฒฝ๋ ๋, asg๋ฅผ scalingํ๋ ๋ฐฉ์์, ํ์ ๋ฉ์์ง ์๊ฐ ํ์์ ๋ฉ์์ง๋ฅผ ์ฒ๋ฆฌํ๋ ์๋ ์ค์ผ์ผ๋ง ๊ทธ๋ฃน์ ํฌ๊ธฐ์ ๋น๋กํ์ฌ ๋ณ๊ฒฝ๋์ง ์๋๋ก ํ๋ ๋ฌธ์ ๊ฐ ๋ฐ์ํ ์ ์๋ค. ์ด๋ฅผ ํด์ํ์๋ฉด, ๋ฉ์์ง ์์ ๋ฐ๋ผ์ asg๊ฐ ์ค์ผ์ผ๋ง๋๋ ๊ฒ์ด ๊ธฐ์ ์ ์ผ๋ก ์๋ฒฝํ ๋น๋กํ๊ธฐ๊ฐ ์ฝ์ง ์์ ๊ฒ ๊ฐ๋ค๊ณ ํ๋จ๋๋ค.
target tracking policy๋ก A backlog per instance metric๋ฅผ SQS metric์ผ๋ก ์ค์ ํ๋ฉด ํด๊ฒฐํ ์ ์๋ค..
NumberOfMessages : 1500
fleet's running capacity : 10 ec2
๊ฐ๋ณ ec2๊ฐ ์ด๋น 100๊ฐ์ message๋ฅผ ์ฒ๋ฆฌํ๋ค๊ณ ๊ฐ์ ํ ๋.
500๊ฐ์ ๋ฉ์์ง๋ฅผ ์ถ๊ฐ์ ์ผ๋ก ์ฒ๋ฆฌํ๊ธฐ ์ํด์, ec2๋ 5๊ฐ๊ฐ ์ถ๊ฐ์ ์ผ๋ก scaling๋๋ค.
#77 AZ ID
์๋ฅผ ๋ค์ด, ํ AWS ๊ณ์ ์ ๊ฐ์ฉ์ฑ ์์ญ us-west-2a๋ ๋ค๋ฅธ AWS ๊ณ์ ์ us-west-2a์ ๋์ผํ ์์น๊ฐ ์๋ ์ ์๋ค.
๋ฐ๋ผ์ ์ ์ํฉ์์ ์๋ฒฝํ ๊ฐ์ az๋ฅผ ์ ์ํ๊ธฐ ์ํด์๋ usw2-az2๊ฐ์ด us-west-2a์ az id๋ฅผ ์ฌ์ฉํด์ผ ํ๋ค.
#78 NAT Gateway vs NAT Instance
NAT Instance๋ง port forwarding, security group, bastion host๋ก ์ฌ์ฉ๊ฐ๋ฅํ๊ณ NAT Gateway๋ ์ด 3๊ฐ ์ ๋ถ ๋ค ๋ถ๊ฐ๋ฅํ๋ค.
#79 You cannot use delay queues to postpone the delivery of only certain messages to the queue by one minute
๋๋ ์ด ํ๋ ์ ์ฒด์ ์ธ ํ์ ์ ์ฉ๊ฐ๋ฅํ์ง๋ง ํน์ ๋ฉ์์ง์๋ง ์ ์ฉํ ์๋ ์๋ค.
ํน์ ๋ฉ์์ง์๋ง ์ ์ฉํ๋ ค๋ฉด, ๋ฉ์์ง ํ์ด๋จธ๋ฅผ ์ด์ฉํด์ผ ํ๋ค.
#80 AWS Cloudtrail vs AWS Config vs AWS Systems Manager
AWS Config : AWS ๋ฆฌ์์ค ๊ตฌ์ฑ ๊ธฐ๋ก ๋ฐ ํ๊ฐ > resource-specific history, audit, and compliance
AWS Config๋ ๊ตฌ์ฑ ๊ธฐ๋ก์ ์ ๊ณตํ๊ธฐ ์ํด AWS ๋ฆฌ์์ค์ ๋ํ ๋ณ๊ฒฝ ์ธ๋ถ ์ ๋ณด๋ฅผ ๊ธฐ๋กํ๋ค. AWS Management ์ฝ์, API ๋๋ CLI๋ฅผ ์ฌ์ฉํ์ฌ ๊ณผ๊ฑฐ ์ด๋ ์์ ์์๋ ๋ฆฌ์์ค ๊ตฌ์ฑ์ด ์ด๋ป๊ฒ ์๊ฒผ๋์ง์ ๋ํ ์ธ๋ถ ์ ๋ณด๋ฅผ ์ป์ ์ ์์ต๋๋ค.
it is a per-region service
AWS Cloudtrail : ์ฌ์ฉ์ ํ๋ ๋ฐ API ์ฌ์ฉ ์ถ์ > account-specific activity and audit
AWS CloudTrail์ ๊ฐ์ฌ, ๋ณด์ ๋ชจ๋ํฐ๋ง ๋ฐ ์ด์ ๋ฌธ์ ํด๊ฒฐ์ ์ง์ํ๋ค. CloudTrail์ AWS ์๋น์ค ์ ๋ฐ์ AWS ์ฌ์ฉ์ ํ๋ ๋ฐ API ์ฌ์ฉ๋์ ์ด๋ฒคํธ๋ก ๊ธฐ๋กํ๋ค. CloudTrail ์ด๋ฒคํธ๋ "๋๊ฐ ๋ฌด์์, ์ด๋์, ์ธ์ ํ์ต๋๊น?"๋ผ๋ ์ง๋ฌธ์ ๋ตํ๋ ๋ฐ ๋์์ด ๋๋ค.
CloudTrail์ ๋ ๊ฐ์ง ์ ํ์ ์ด๋ฒคํธ๋ฅผ ๊ธฐ๋กํฉ๋๋ค. S3 ๋ฒํท ์์ฑ ๋๋ ์ญ์ ์ ๊ฐ์ ๋ฆฌ์์ค์ ๋ํ ์ ์ด ํ๋ ์ธ ์์ ์ ์บก์ฒํ๋ ๊ด๋ฆฌ ์ด๋ฒคํธ
S3 ๊ฐ์ฒด ์ฝ๊ธฐ ๋๋ ์ฐ๊ธฐ์ ๊ฐ์ ๋ฆฌ์์ค ๋ด ๋ฐ์ดํฐ ํ๋ ์ธ ์์ ์ ์บก์ฒํ๋ ๋ฐ์ดํฐ ์ด๋ฒคํธ
A trail can be applied to All Regions(default) or a single Region
AWS Systems Manager : AWS ๋ฐ ์จํ๋ ๋ฏธ์ค ๋ฆฌ์์ค์ ๋ํ ์ด์ ์ธ์ฌ์ดํธ ํ๋ณด : ๋ฆฌ์์ค ๊ทธ๋ฃน, ์ค์์ง์คํ, aws๋ฆฌ์์ค ๊ด๋ฆฌ์ ์ค์ฌํ
#81 AWS Transfer Family
AWS Transfer Family๋ SFTP, FTPS ๋ฐ FTP๋ฅผ ํตํด Amazon S3 ๋ฐ Amazon EFS ์ํ์ผ๋ก ์ง์ ํ์ผ์ ์ ์กํ ์ ์๋๋ก ์์ ๊ด๋ฆฌํ ์ง์์ ์ ๊ณตํ๋ค.
๋ฐ๋ณต์ ์ธ ๋น์ฆ๋์ค ๊ฐ ํ์ผ ์ ์ก์ ์ฌ์ฉ
Windows ํ์ผ ์๋ฒ์ฉ Amazon FSX๋ ์ง์ํ์ง ์๋๋ค.
#82 AWS Storage Gateway
์จํ๋ ๋ฏธ์ค์์ ๋ฌด์ ํ์ ํด๋ผ์ฐ๋ ์คํ ๋ฆฌ์ง์ ์ก์ธ์คํ ์ ์๊ฒ ํด์ฃผ๋ ํ์ด๋ธ๋ฆฌ๋ ํด๋ผ์ฐ๋ ์คํ ๋ฆฌ์ง ์๋น์ค
s3, fsx for windows file server์ ์ ๊ทผ, ๋ฐฑ์ ์ ํด๋ผ์ฐ๋๋ก ์ด๋ํ๊ณ , ํด๋ผ์ฐ๋ ์คํ ๋ฆฌ์ง์์ ์ง์๋๋ ์จํ๋ ๋ฏธ์ค ํ์ผ ๊ณต์ ๋ฅผ ์ฌ์ฉ๋ ํฌํจ
#83 AMI
You can copy both Amazon EBS-backed AMIs and instance-store-backed AMIs.
You can share an AMI with another AWS account
To copy an AMI that was shared with you from another account, the owner of the source AMI must grant you read permissions for the storage that backs the AMI, either the associated EBS snapshot (for an Amazon EBS-backed AMI) or an associated S3 bucket (for an instance store-backed AMI).
Copying an AMI backed by an encrypted snapshot cannot result in an unencrypted target snapshot
์๋๋ ami copy ์๋๋ฆฌ์ค์ด๋ค. Amazon EBS-backed AMI์ ๋ํด์๋ง ์ ์ฉ๋๊ณ , instance store-backed AMI๋ encryption status์๋ง ์ ์ฉ๋๊ธฐ ๋๋ฌธ์ encrypted status๋ฅผ ๋ฐ๊ฟ ์ ์๋ค.
1
Unencrypted-to-unencrypted
Yes
2
Encrypted-to-encrypted
Yes
3
Unencrypted-to-encrypted
Yes
4
Encrypted-to-unencrypted
No
#84 Tenancy of instance
You can change the tenancy of an instance from dedicated to host
You can change the tenancy of an instance from host to dedicated
dedicated์ default, host์ default ๊ฐ์ ๋ณ๊ฒฝ์ ๋ถ๊ฐ๋ฅํ๋ค.
#85 ์ํคํ ์ฒ
๋ง์ดํฌ๋ก์๋น์ค ์ค ์ด๋ค ์๋น์ค๋ ๋น ๋ฅด๊ฒ ์คํ๋๊ณ , ์ด๋ค ์๋น์ค๋ ๋๋ฆฌ๊ฒ ์คํ๋๋ฉด decoupling์ ๊ฒํ ํด์ผ ํ๋ค.
#86 Cloudhub
Vpc์ virtual private gateway๋ฅผ ๋๊ณ , vpcํ๊ฒฝ๊ณผ on-premiseํ๊ฒฝ์ ์ฐ๊ฒฐํ๋ค.
hub and spoke๋ชจ๋ธ๋ก vpc์ ์จํ๋ ๋ฏธ์ค ๋ผ๋ฆฌ ์์ ๋กญ๊ฒ ์ฐ๊ฒฐํ ์ ์๋ค.
#87
A recovered instance is identical to the original instance, including the instance ID, private IP addresses, Elastic IP addresses, and all instance metadata : ๋ณต๊ตฌ๋ ์ธ์คํด์ค๋ ๊ธฐ์กด ์ธ์คํด์ค์ instance ID, private IP addresses, Elastic IP addresses, and all instance metadata๊ฐ ๊ฐ๋ค.
If your instance has a public IPv4 address, it retains the public IPv4 address after recovery
#88 Amazon EC2 Auto Scaling chooses the policy that provides the largest capacity : ๋ ์ ์ฑ ์ด ์ถฉ๋ํ๋ฉด ๊ฐ์ฅ ํฐ capacity๋ฅผ ์ฐ์ ํด์ ๋์ํ๋ค.
#89 Step Function vs Simple WorkFlow Service
Step Function : JSON์ผ๋ก ์ํ ์์คํ ์ ์ ์ํ๋ค.
Simple WorkFlow Service : ํ๋ก๊ทธ๋๋ฐ ์ธ์ด๋ก Decider ํ๋ก๊ทธ๋จ์ ์์ฑํ๊ฑฐ๋ Flow Framework๋ฅผ ํตํด ๋๊ธฐ์ ์ํธ ์์ฉ์ ๊ตฌ์ฑํ๋ ํ๋ก๊ทธ๋๋ฐ ๊ตฌ๋ฌธ์ ์์ฑ
#90 ์ํคํ ์ฒ
With a sharp increase in the number of users, the system has become slow and sometimes even unresponsive as it does not have a retry mechanism
์ฌ์ฉ์ ์๊ฐ ๊ธ์ฆํ๋ฉด์ ์ฌ์๋ ๋ฉ์ปค๋์ฆ์ด ์์ด, ์์คํ ์ด ๋๋ ค์ง๊ณ ๋๋ก๋ ๋ฐ์์ด ์๋ ๊ฒฝ์ฐ๋ ์์๋ค.
ํด๊ฒฐ์ฑ : Use Amazon Kinesis Data Streams to ingest the data, process it using AWS Lambda or run analytics using Kinesis Data Analytics
#91
Amazon EFS uses the Network File System protocol. EFS does not support SMB protocol.
Amazon FSx for Windows File Server, File Gateway Configuration of AWS Storage Gateway support SMB Protocol.
#92 Spot Instance
If the request is persistent and you stop your Spot Instance, the request only opens after you start your Spot Instance.
If a spot request is persistent, then it is opened again after your Spot Instance is interrupted
#93 Use Amazon GuardDuty to monitor any malicious activity on data stored in S3. Use Amazon Macie to identify any sensitive data stored on S3
#94 SCP(Service Control Policies)
SCP must have an explicit Allow (does not allow anything by default)
Does not apply to the Master Account
SCP is applied to all the Users and Roles of the Account, including Root user
Master Account์๋ ์ ์ฉ์ด ๋์ง ์์ง๋ง, ๋ฃจํธ ์ ์ ๋ ์ ์ฉ์ด ๋๋ค๋ ์ ์ฃผ์
The SCP does not affect service-linked roles : SCP์ service-linked roles๋ ๊ด๋ จ์ด ์๋ค.
#95 ์๋ฒ๋ฆฌ์ค ์ํคํ ์ฒ
Host the static content on Amazon S3 and use Lambda with DynamoDB for the serverless web application that handles dynamic content. Amazon CloudFront will sit in front of Lambda for distribution across diverse regions
#96 Weekly Job for 5 minutes์ด ํ์ํ ๊ฒฝ์ฐ ์ฌ์ฉ๊ฐ๋ฅํ ์ํคํ ์ฒ
Schedule a weekly CloudWatch event cron expression to invoke a Lambda function that runs the database rollover job
#97 Route 53 alias vs cname
You should also note that Route 53 doesn't charge for alias queries to AWS resources but Route 53 does charge for CNAME queries
Additionally, an alias record can only redirect queries to selected AWS resources such as S3 buckets, CloudFront distributions, and another record in the same Route 53 hosted zone. : alias record๋ aws resource๋ง์ ๋์์ผ๋ก ํ๋ค.
However a CNAME record can redirect DNS queries to any DNS record. So, you can create a CNAME record that redirects queries from app.covid19survey.com to app.covid19survey.net.
#98 Internet Gateway
An Internet Gateway serves two purposes
provide a target in your VPC route tables for internet-routable traffic
perform network address translation(NAT) for instances that have been assigned public IPv4 addresses
#99 NLB
Traffic is routed to instances using the primary private IP address specified in the primary network interface for the instance
#100 DMS ์ฌ์ฉ ์์
Use AWS Database Migration Service to replicate the data from the databases into Amazon Redshift
#101 Elastic Fabric Adapter - HPC
An Elastic Fabric Adapter(EFA) is a network device that you can attach to your Amazon EC2 instance to accelerate High Performance Computing (HPC) and machine learning applications
#102 VPC
Using VPC sharing, an account that owns the VPC(owner) shares one or more subnets with other accounts(participants) that belong to the same organization from AWS Organizations. The owner account cannot share the VPC itself. : ์ค๋ ๊ณ์ ์ VPC ์์ฒด๋ฅผ ๊ณต์ ํ ์ ์๋ค.
Use VPC sharing to share one or more subnets with other AWS accounts belonging to the same parent organization from AWS Organizations : VPC Sharing์ vpc ์์ฒด๋ฅผ ๊ณต์ ํ๋ ๊ฒ์ด ์๋๋ผ vpc ์๋ธ๋ท์ ๊ณต์ ํ๋ ๊ฒ์ด๋ค.
๋ํ VPC sharing์ owner ๊ณ์ ์์ subnet์ ๊ด๋ฆฌํ๊ณ ์ด ์๋ธ๋ท์ ๊ณต์ ํ๋ ๊ฒ์ด๊ธฐ ๋๋ฌธ์ ์ค์๊ด๋ฆฌ๊ฐ ๋๋ค๋ ์ฅ์ ์ด ์๋ค.
#103 Global Accelerator
AWS Global Accelerator is a networking service that helps you improve the availability and performance of the applications that you offer to your global users. : ์ ํ๋ฆฌ์ผ์ด์ ์ ๋คํธ์ํฌ ์ฑ๋ฅ ํฅ์์ ํจ๊ณผ์ ์
It provides static IP addresses that provide a fixed entry point to your applications and eliminate the complexity of managing specific IP addresses for different AWS Regions and Availability Zones.
AWS Global Accelerator always routes user traffic to the optimal endpoint.
Global Accelerator is a good fit for non-HTTP use cases, such as gaming(UDP), IoT(MQTT), or Voice over IP.
#104 CloudFront
CloudFront supports HTTP/RTMP protocol based requests.
CloudFront do not support UDP.
CloudFront points of presence(POPs) : CloudFront also has regional edge caches that bring more of your content closer to your viewers, even when the content is not popular enough to stay at a POP, to help improve performance for that content.
Regional edge caches help with all types of content, particularly content that tends to become less popular over time. : ์์ฃผ ์ ๊ทผ๋์ง ์๋ static ์์์ ๋ํด์๋ ์ง์ํ๋ค.
Examples include user-generated content, such as video, photos, or artwork, e-commerce assets such as product photos and videos; and news and event-related content that might suddenly find new popularity. : ์ ์ ์ ์ ๋น๋์ค, ์ฌ์ง, ์ํธ์ํฌ ๋ฑ ๋ชจ๋ ์ง์์ Cloudfront์์ ๋ชจ๋ ์ง์
Cloudfront๋ 1GB ๋ฏธ๋ง์ธ static ์์์ ์บ์ํ๊ธฐ ์ ํฉํ๋ค. 1GB ์ด์์ธ ์์์ ๋ํด์๋ S3 Transfer Acceleration(Cloudfront์ ๋ง์ฐฌ๊ฐ์ง๋ก ๊ธ๋ก๋ฒํ ์๋น์ค์ด๊ธฐ ๋๋ฌธ์ ๊ธ๋ก๋ฒํ ์ ํ๋ฆฌ์ผ์ด์ ์ ์ ํฉ)์ ์ฌ์ฉํ๋ ๊ฒ์ด ์ข๋ค.
S3 Transfer Acceleration improves transfer performance by routing traffic through Amazon CloudFrontโs globally distributed Edge Locations and over AWS backbone networks, and by using network protocol optimizations.
#105 AWS Managed Microsoft AD vs AD Connector vs Simple AD
AWS Managed Microsoft AD : AWS Managed Microsoft AD would also allow you to run directory-aware workloads in the AWS Cloud. AWS Managed Microsoft AD is your best choice if you have more than 5,000 users and need a trust relationship set up between an AWS hosted directory and your on-premises directories. : directory-aware workloads, trust relationships with other domains์ด simple ad์ ad connector์ ๋น๊ตํ์ ๋ AWS Managed Microsoft AD๋ง ๊ฐ์ง๊ณ ์๋ ํน์ฑ์ด๋ค.
AD Connector : Just remember that you should use AD Connector if you only need to allow your on-premises users to log in to AWS applications with their Active Directory credentials
Simple AD : Simple AD is the least expensive option and your best choice if you have 5,000 or fewer users and donโt need the more advanced Microsoft Active Directory features such as trust relationships with other domains.
#106 RDS Read Replica
Serving read traffic while the source DB instance is unavailable. : ํ์ง๋ง ๋ง์คํฐ db๊ฐ unavailableํด์ง๋ฉด read replicat์ ๋ฐ์ดํฐ๋ ๋๊ฒฐ ์ํ๊ฐ ๋๋ค.
You may use a read replica for disaster recovery of the source DB instance, either in the same AWS Region or in another Region. : read replica๋ฅผ ๊ธ๋ก๋ฒํ disaster recovery์ ์ฌ์ฉํ ์ ์๋ค.
#107 aws x-ray
trace and debug
Last updated