AWS pt 0-1 Flashcards
1
Q
RDS (Multi-AZ)
A
- High availability
- Non-Aurora: synchronus replication
- Aurora: asynchronous replication
- Always span at least two AZs within a single region
2
Q
RDS (Multi-Region Deployments)
A
- Main purpose is disaster recovery and local performance
- Asynchronous replication
- Each region can have a Multi-AZ deployment
3
Q
Read Replicas
A
- Main purpose is scalability
- Asynchronus replication
- Can be within an Availability zone, Cross-AZ, or Cross-Region
- Up to 15
- Writing is limited to the main DB
4
Q
DynamoDB
A
- key-value and document database delivers sub-millisecond performance at any scale
- Serverless: No need for hardware provisioning, setup and configuration, throughput capacity planning, replication, sw patching or cluster scaling
- integrated with IAM
-Standard and infrequent Access
5
Q
Security Shared Responsibility (customer)
A
- Security βinβ the cloud
- Platform, app, IAM
- OS, Network and firewall config
- Client side data, encrytption, data integrit yand authentication
- Server-side encryption
- Network traffic protection
5
Q
Storage Gateway
A
- File Gateway (store files as objects in S3)
- Tape Gateway
- Volume Gateway
6
Q
Security shared responsibility (AWS)
A
- SW: compute, storage, db, networking
- HW/Global infra: Regions, AZ, edge lcoations
7
Q
RDS
A
- (OLTP)
- Managed DB: cannot ssh into your instance
- Automated provisioning, OS patching
- Auto-backups = Restoring to a specific timestamp
- Storage backed by EBS
- Scalable
- Deployments (Create a snapshot of your DB -> Transfer into a different AZ)
8
Q
Amazon Aurora
A
- SQL
- Storage auto grows in increments from 10 GB to 128 TB
- Costs more than RDS
- Automated Aurora Serverless:
- DB instantiation & auto-scaling based on usage
- SQL work in Aurora serverless
-No capacity planning/ Least managemet overhead - Pay per second
- Good for infrequent or unpredictable workloads
9
Q
Amazon Elasticache
A
- Get managed Redis or Memcached
- Load off db for intensive workloads -> lower latency = high performance
- EC2 will read and write from cache if faster than RDS
10
Q
Redshift
A
- Online analytical processing
- Load every hour
- Highly available: Massively parallel query execution
- Pay as you go: Based on instance provisional
- SQL interface
- Redshift Serverless:
- Auto provisions + scales Data warehousing
- Pay for what you use
- Use cases: Reporting dashboards, real time analysis
11
Q
Amazon EMR
A
- Hadoop Clusters (Big data) analysis: Use hundreds of EC2 instances (Apache Spark)
- Provisions and configures these EC2 instances
- Auto-scaling and integrated with spot instances
- Data processing, ML, web-indexing and big data
12
Q
Amazon Athena
A
- Analyze data in S3 using Serverless SQL
- BI, analytics, reporting, ELB logs, cloudtrail Logs
13
Q
Amazon Quicksight
A
- serverless, ML powered for interactive dashboards
- Fast, auto scalable, embeddable, with pre-session pricing
- Building visualizations, ad-hoc analysis, business insight with data
14
Q
DocumentDB
A
- MongoDB
- Json data
- Fully managed, highly available, replication across 3 AZs
- Storage auto-grows in increments of 10 GB
- Auto scales to millions of requests