cloud-platforms
Cloud Platforms
AWS Services and Best Practices
Compute Services
-
EC2 (Elastic Compute Cloud): Virtual servers in the cloud
- Use instance types appropriate for workload requirements
- Implement Auto Scaling Groups for elasticity
- Use Spot Instances for fault-tolerant, interruptible workloads
- Leverage EC2 Fleet for diverse instance strategies
-
Lambda: Serverless compute service
- Ideal for event-driven architectures
- Use for short-lived, stateless functions
- Implement dead-letter queues for failed invocations
- Monitor with CloudWatch metrics and logs
-
ECS (Elastic Container Service): Container orchestration
- Use Fargate for serverless container execution
- Implement task definitions with resource limits
- Use service auto-scaling based on metrics
- Configure load balancing with ALB/NLB
-
EKS (Elastic Kubernetes Service): Managed Kubernetes
- Use managed node groups for simplified operations
- Implement pod autoscaling (HPA, VPA)
- Use AWS VPC CNI for networking
- Integrate with IAM for service accounts
Storage Services
-
S3 (Simple Storage Service): Object storage
- Use lifecycle policies for cost optimization
- Implement versioning for data protection
- Use S3 Transfer Acceleration for faster uploads
- Configure CORS for cross-origin access
- Enable S3 Event Notifications for automation
-
EBS (Elastic Block Store): Block storage
- Choose volume type based on workload (gp3, io2, etc.)
- Use multi-attach for high availability
- Implement snapshots for backup
- Monitor volume metrics for performance
Database Services
-
RDS (Relational Database Service): Managed relational databases
- Use Multi-AZ deployments for high availability
- Enable read replicas for scaling reads
- Use automated backups and point-in-time recovery
- Implement parameter groups for configuration
-
DynamoDB: NoSQL database
- Design partition keys for even distribution
- Use on-demand mode for unpredictable workloads
- Implement TTL for automatic data expiration
- Use DynamoDB Accelerator (DAX) for caching
Infrastructure as Code
- CloudFormation: AWS native IaC
- Use stacks for resource organization
- Implement nested stacks for modularity
- Use change sets for safe updates
- Leverage CloudFormation exports for cross-stack references
Networking
- VPC (Virtual Private Cloud): Isolated network environment
- Use public and private subnets for tiered architecture
- Implement NAT Gateways for private subnet outbound access
- Use VPC endpoints for private connectivity to AWS services
- Configure route tables and security groups properly
Azure Services and Best Practices
Compute Services
-
Azure Virtual Machines: Virtual servers
- Use managed disks for storage
- Implement availability sets for high availability
- Use Azure Spot VMs for cost savings
- Configure extensions for monitoring and management
-
Azure Functions: Serverless compute
- Use Consumption plan for event-driven workloads
- Implement Durable Functions for stateful orchestrations
- Use Application Insights for monitoring
- Configure function app scaling
-
Azure Kubernetes Service (AKS): Managed Kubernetes
- Use Azure CNI for advanced networking
- Implement cluster autoscaler
- Use Azure AD integration for authentication
- Configure pod identity for secure access to Azure resources
Storage Services
-
Azure Blob Storage: Object storage
- Use access tiers (Hot, Cool, Archive) for cost optimization
- Implement lifecycle management policies
- Use blob versioning for data protection
- Configure CORS and shared access signatures
-
Azure Disk Storage: Block storage
- Choose disk type based on workload (Premium SSD, Ultra Disk)
- Use Azure Disk Encryption for data at rest
- Implement snapshots for backup
- Monitor disk performance metrics
Database Services
-
Azure SQL Database: Managed SQL database
- Use vCore-based or DTU-based purchasing models
- Implement geo-replication for disaster recovery
- Use transparent data encryption
- Configure automatic backups
-
Azure Cosmos DB: Globally distributed NoSQL database
- Choose appropriate API (SQL, MongoDB, Cassandra, etc.)
- Use multi-master replication for global availability
- Implement consistency levels based on requirements
- Use throughput provisioning with RU/s
Infrastructure as Code
- Azure Resource Manager (ARM) Templates: Azure native IaC
- Use parameter files for environment-specific configurations
- Implement linked templates for modularity
- Use deployment scripts for post-deployment actions
- Leverage template specs for reusability
Networking
- Azure Virtual Network (VNet): Isolated network
- Use subnets for network segmentation
- Implement NSGs for security rules
- Use Azure Firewall for network protection
- Configure VNet peering for connectivity
GCP Services and Best Practices
Compute Services
-
Compute Engine: Virtual machines
- Use custom machine types for optimized workloads
- Implement instance groups for auto-scaling
- Use preemptible VMs for cost savings
- Configure startup and shutdown scripts
-
Cloud Functions: Serverless compute
- Use 2nd generation functions for better performance
- Implement event triggers for automation
- Use Cloud Logging and Cloud Monitoring
- Configure function deployment settings
-
Google Kubernetes Engine (GKE): Managed Kubernetes
- Use Autopilot for fully managed clusters
- Implement node auto-provisioning
- Use Workload Identity for secure access
- Configure network policies for pod security
Storage Services
-
Cloud Storage: Object storage
- Use storage classes (Standard, Nearline, Coldline, Archive)
- Implement lifecycle management rules
- Use object versioning for data protection
- Configure signed URLs and ACLs
-
Persistent Disks: Block storage
- Choose disk type (Standard, Balanced, Extreme)
- Use regional disks for high availability
- Implement snapshots for backup
- Monitor disk I/O and throughput
Database Services
-
Cloud SQL: Managed relational databases
- Use high availability configuration
- Implement read replicas for scaling
- Use automated backups and point-in-time recovery
- Configure SSL/TLS connections
-
Cloud Spanner: Globally distributed SQL database
- Use multi-region configuration for global availability
- Design schema for optimal performance
- Implement instance sizing and scaling
- Use database roles for access control
Infrastructure as Code
- Deployment Manager: GCP native IaC
- Use Jinja or Python templates
- Implement composite types for reusability
- Use deployment manifests for organization
- Configure preview deployments
Networking
- Virtual Private Cloud (VPC): Isolated network
- Use subnets for network segmentation
- Implement VPC peering for connectivity
- Use Cloud NAT for private subnet outbound access
- Configure firewall rules for security
Multi-Cloud Strategies and Considerations
Multi-Cloud Approaches
- Multi-Cloud for Resilience: Distribute workloads across providers for disaster recovery
- Best-of-Breed Services: Use specific services from each provider based on strengths
- Vendor Lockout Mitigation: Avoid single-provider dependencies
- Cost Optimization: Leverage competitive pricing and spot markets
Multi-Cloud Challenges
- Complexity: Increased operational complexity and management overhead
- Consistency: Maintaining consistency across different platforms
- Networking: Cross-cloud connectivity and latency considerations
- Identity and Access Management: Unified identity across providers
Multi-Cloud Best Practices
- Abstraction Layers: Use abstraction layers (Terraform, Pulumi) for multi-cloud deployments
- Standardization: Standardize on common tools and practices
- Observability: Implement unified monitoring and logging across clouds
- Security: Implement consistent security policies across all platforms
More from davincidreams/agent-team-plugins
blender
Blender interface, workflows, and 3D production pipeline
220rigging
Rigging fundamentals, skeleton setup, and animation controls
16animation
Animation principles, techniques, and best practices for 3D animation
13vroid
Vroid Studio, VRM format, and VTuber avatar creation
10technical-writing
Technical writing principles and best practices for creating clear, accurate documentation
9unreal
Unreal Engine patterns, Actor/Component model, Blueprints vs C++, and best practices
8