Cloud Architecture (Unit 5)

Cloud Architecture - CA - Unit 5 -SRMIST Syllabus - Short Notes

Cloud Architectural Patterns

• Scalability - Number of users who can effectively use a service at the same time.
• Horizontal Scaling - Scaling on demand, by adding new nodes.
  
• Vertical Scaling - Scaling on demand, by adding resources to existing nodes.
  
  
  
• Queue-Centric - First come first serve.
  
• Auto-Scaling - Automatically scales upon demand in a day.
  
• Eventual Consistency Primer - Making the cloud database consistent.
  
  
  
• Map Reduce - Like Hadoop, uses 2 functions one to map data to functions and the next to reduce the workload of the processor by distributing data to multiple functions.
  
  
  
• Database Sharding - Effective use of database storage by distributing data to multiple nodes.
  
  
  
• Node Failure - Recovery or response to node failure.
  
  
  
• Network Latency Primer - Reduce downtime and delay in communication over a network.
• Move applications closer to users.
• Move data closer to users.
• Ensure nodes & applications are close by.
• CDN - Using temporary storage to retrieve data much more quickly.
  
  
  
• Multisite deployment - Multiple data center deployments to reduce latency.
  
  
  

Network Connectivity

• TOR - Less cabling.
  
  
  
• EOR - Too much cabling.
  
  
  
• Standalone Server - A server that runs all on its own, authentication, security services, etc.
  
  
  
• Blade Server - Modular server with a self-cooling system.
  
  
  
• Stateless Server - No state, user information disregarded.
  
  
  
• Server Cluster - Group of servers that work simultaneously.