AWS FAQs Flashcards
EC2 - What monthly uptime is guaranteed?
99.99% (EC2 & EBS)
EC2 - What are A1 instances?
General purpose instane with AWS Graviton Processors
EC2 - What are A1 instances used for?
ARM based instances, that deliver good performance, cost savings with a small available memory footprint
EC2 - Use cases for M5 instances
Higher memory footsprint for intensive Web, mobile, gaming or analytics applications
EC2 - What is burstable performance?
Instance provides a baseline level of performance, but can scale up.
When going over the baseline it consumes CPU credits
EC2 - What are High Memory instances usedd for?
Enterprise Software with large RAM use-cases, SAP e.g.
EC2 - When should a memory-optimized instance be used?
All in-memory applications, like databse or analytics
EBS - What happens to the data when the system terminates?
Unlike local instance storage, EBS storage is persisted and not lost
EBS - What are the differences between the volumes types?
SSD: Transactional workloads based on high IOPS
HDD: Throughput based on MB/s
EBS - Can volumes and snapshots be encrypted?
Yes, both for volumes and snapshots. This enables the user to meet security and compliance requirements
Elastic IP - Explain the limits of IP addresses and the use charges
IPv4 IPs are limited, therefore AWS is commited to keep the use efficiently.
So reserving and IP and not attaching it to an instance will be charged.
An public is not required for all instances
ELB - Difference between Classic and Application Load Balancer
Classic: simple load balacer to target EC2 instances
Application: adanced routing for container based architecture or microservices
Can API calls be tracked?
Yes, with CloudTrail
EC2 - Why should an instance be hibernated?
Hibernated instances can be activated quickly.
Content from memory are peristed to an EBS volume
EC2 - How is EC2 billed?
- Paid by use by the hour or second
- Billings starts and the boot sequence and ends on when the instance is terminated
EC2 - How is traffic beteen AZs or regions billed?
1) Data out to another AWS region + 2) Data in from another region
EC2 - How are AZs isolated from each other?
Physically dinstint with independent infrastructure.
Common points of failure (generator / cooling) is not shared across AZs
EC2 - What are the five instance type families?
1) General Purpose
2) Compute Optimized
3) Memory Optimized
4) Storage Optimized
5) Accelerated Computing
VPC - What is AWS VPC?
Virtual Networking environment, with IP adress ranges, subnets, route tables and network gateways.
COnfigured correctly provides public facing subnets facing the customer and private subnets protecting the data.
VPC - What are the components of VPC?
- Subnet: segment of ip ranges in isolated groups
- Internet Gateway: connection to the internet
- NAT Gateway: Access for private subnets to the internet
- Virtual Private Gateway: internal VPN in AWS
- Peering Connection: connectio between VPCs
- VPC Endpoints: connects AWS services (S3 and Dynam DB)without VPN or gateways
VPC - How can an EC2 instance be secured in a VPC?
Security Groups: Allows ports and IPs - per default everything is disallowed (only allow)
Access Control Lists: Handles network traffic for each subnet (has allow/deny rules)
VPC - can network traffic be monitored?
Yes, with Amazon VPC Flow Logs
VPC - Can regional VPCs be connected?
Interregional VPC peering is available (except with china)
S3 - How much data can be stored?
- Number is unlimited
- Size between 0 bytes and 5 TB
- SIngle PUT up to 5GB, recommended to upload >100mb files in multipart