AWS Architecture and Services

• AWS has 20 regions globally, with 5 more regions coming soon, and each region has 2 or more availability zones (AZs) for high availability of services. • AWS global data centers have edge locations, which are like caching devices across 100+ cities worldwide, to improve performance by reducing latency. • An AWS region is a geographic area, and each region consists of 2 or more AZs, which are used for high availability of applications. • When designing an architecture, it's typical to keep machines in different AZs so that if one AZ goes down, the application remains available. • An AWS account is a top-level entity that allows deployment of infrastructure in any AWS region, and each region has 2 or more AZs. • AWS services have different scopes, with some working at the account level (e.g., billing, IAM), some at the region level (e.g., S3, DynamoDB), and some at the AZ level (e.g., EC2, RDS, EBS). • There are 130+ AWS services, which can be broadly categorized into computing power, analytics, storage, database, network, management, application, and development services. • AWS services can be used to build an application architecture, such as a social media application, and can be categorized into different tiers (e.g., web server, application server, database). • A three-tier architecture typically consists of a web server, application server, and database, and can be scaled vertically or horizontally to handle increased load. • Load balancers can be used to distribute load across multiple web servers and application servers, and DNS services can be used to map domain names to load balancer IP addresses. • Relational databases may not be suitable for large-scale data storage, and NoSQL databases may be more suitable for handling large amounts of data and connections.Here are the detailed bullet points summarizing the text:

• Facebook's architecture involves storing data in both relational and NoSQL databases, with relational databases potentially becoming a performance bottleneck.

• To optimize performance, database caches like Redis or Memcached are used to store frequently accessed data, reducing the load on relational databases.

• External storage solutions like file storage or shared filesystems are used to store media files, such as pictures and videos, due to the size limitations of disks attached to VMs.

• Content filters are used to analyze and filter out objectionable content from uploaded media files.

• Clickstream analysis is used to capture and analyze user behavior, such as clicks and likes, in real-time.

• A Hadoop platform is used to perform computing on distributed systems, and a data warehouse is used to store and analyze data for business intelligence.

• A business intelligence tool is used to query and analyze data, generating reports that inform business decisions.

• The architecture also includes a notification service, email service, and messaging queue service, such as RabbitMQ or JMS queues.

• A monitoring service is used to continuously monitor the health of the application and infrastructure.

• To deploy this architecture on AWS, various services are used, including VPC, EC2, EBS, RDS, DynamoDB, Elasticache, S3, Rekognition, Lambda, Kinesis, EMR, and Redshift.

• Auto scaling is used to scale EC2 instances horizontally based on incoming traffic.

• A content delivery network (CDN) like CloudFront is used to cache and distribute static content across edge locations.

• SNS and SES are used for sending messages and bulk emails, respectively.

• SQS is used for messaging queues, and CloudWatch is used for monitoring and alarms.

• API Gateway is used to manage REST APIs, and Cognito is used for user management and identity management.

• IAM is used for managing access and authentication, and KMS is used for encrypting data.

• ACM is used for managing digital certificates, and WAF is used for application-level firewalls.

• Inspector is used for compliance and patch management, scanning machines for vulnerabilities and generating reports.

• The architecture also includes development and DevOps services, such as CodeCommit, CodeBuild, CodePipeline, and CodeDeploy.- Manual deployment of AWS services can take around 2-3 days without errors, but with AWS, infrastructure can be coded and deployed within 30 minutes using CloudFormation.

• CloudFormation is a powerful service that provisions infrastructure from scratch using a JSON or YAML template.
• DevOps teams write CloudFormation templates, which are then used by the CloudFormation service to create infrastructure.
• AWS CodeCommit is a code repository service that allows developers, DevOps, and QA teams to check-in code.
• CloudFormation templates are written in JSON or YAML code and are checked into CodeCommit.
• AWS CodeBuild takes source code, builds it using build tools like Ant or Maven, and performs unit tests to produce artifacts.
• Artifacts are application executables, such as EXEs or binaries.
• CodeBuild produces artifacts, which are then deployed to EC2 machines using CodeDeploy.
• CodeDeploy is a deployment service that deploys artifacts to application servers running in EC2.
• CodePipeline is a CI/CD pipeline service that automates the build, test, and deployment process.
• CodePipeline integrates with CodeCommit, CodeBuild, and CodeDeploy to automate the entire development and deployment process.
• AWS CodeStar integrates with project management tools like JIRA, providing complete SDLC control and development speed management.
• AWS services mentioned include CloudFormation, CodeCommit, CodeBuild, CodeDeploy, CodePipeline, and CodeStar.