Cloud computing is the delivery of computing services over the internet. These services include:
- Common IT infrastructure (virtual machines, storage, databases, networking)
- Advanced services (IoT, machine learning, artificial intelligence)
Unlike traditional datacenters, cloud computing isn't constrained by physical infrastructure, allowing for rapid expansion of IT capabilities when needed.
The responsibility for security and management is shared between the cloud provider and the consumer:
Cloud Provider is Always Responsible For:
- Physical datacenter
- Physical network
- Physical hosts
Customer is Always Responsible For:
- Information and data stored in the cloud
- Devices allowed to connect to your cloud
- Accounts and identities of people, services, and devices
Responsibility Based on Service Model:
- Operating systems
- Network controls
- Applications
- Identity and infrastructure
A private cloud delivers IT services over the internet for a single entity. It provides greater control but comes with higher costs and fewer benefits compared to public cloud. It may be hosted:
- In your on-site datacenter
- In a dedicated offsite datacenter (potentially by a third party)
Built, controlled, and maintained by a third-party cloud provider. Anyone can purchase and access resources. The general public availability is the key difference from private clouds.
A computing environment using both public and private clouds in an interconnected way. Benefits include:
- Handling increased temporary demand by deploying public cloud resources
- Providing additional security layers
- Flexibility in choosing which services to keep in public vs private cloud
Using multiple public cloud providers simultaneously. This might occur when:
- Using different features from different providers
- Migrating from one provider to another
Example: Hybrid Multi-Cloud for Enterprises A large enterprise combining multiple cloud platforms:
- AWS: Hosting main application backend
- Azure: Used for Microsoft-based workloads (Active Directory, Office 365)
- Google Cloud: Handling AI/ML workloads with TensorFlow and BigQuery
Benefit: Avoid vendor lock-in and optimize cloud resources
- Azure Arc: A set of technologies that helps manage your cloud environment across public, private, hybrid, or multi-cloud configurations
- Azure VMware Solution (AVS): Provides a seamless way to run VMware workloads in Microsoft Azure without significant changes
One-time, up-front expenditure to purchase or secure tangible resources (e.g., new building, datacenter, company vehicle)
Spending money on services or products over time (e.g., renting convention center, leasing vehicles, cloud services)
Cloud computing uses a consumption-based OpEx model with benefits:
- No upfront costs
- No need to purchase and manage potentially underutilized infrastructure
- Pay for more resources only when needed
- Stop paying for resources when no longer needed
A formal agreement between a customer and service provider guaranteeing stated services. Azure SLAs are represented as percentages, with service availability measured as uptime.
The ability to adjust resources to meet demand, ensuring you're not overpaying for unused services.
Types of Scaling:
- Vertical Scaling: Increasing or decreasing the capabilities of resources (adding CPU/RAM)
- Horizontal Scaling: Adding or removing number of resources (like VMs or containers)
The ability of a system to recover from failures and continue functioning. The cloud's decentralized design naturally supports reliable and recoverable infrastructure.
Performance Predictability: Focuses on predicting resources needed for positive customer experiences, using:
- Autoscaling
- Load balancing
- High availability
Cost Predictability: Focuses on forecasting cloud spending using tools like Total Cost of Ownership (TCO) calculator or Pricing Calculator.
Ensuring cloud resources follow company policies and regulatory rules:
- Templates for consistency
- Automatic updates to standards
- Auditing & compliance monitoring
- Mitigation strategies
Benefits of early governance:
- Secure cloud environment
- Compliance with regulations
- Automated software updates
- Reduced security risks and downtime
Methods for managing cloud resources:
- Web portal
- Command line interface
- APIs
- PowerShell
Cloud management capabilities:
- Automatic scaling based on need
- Template-based resource deployment
- Resource health monitoring
- Automatic alerts based on metrics
The most flexible cloud service category with maximum control. The provider maintains hardware, network connectivity, and physical security.
Common Use Cases:
- Lift-and-shift migration
- Testing and development environments
Middle ground between IaaS and SaaS. The provider maintains physical infrastructure, internet access, operating systems, databases, and development tools.
Common Use Cases:
- Development frameworks
- Analytics and business intelligence
The most complete cloud service model. You rent or use fully developed applications with minimal technical expertise required.
Common Use Cases:
- Email and messaging
- Business productivity applications
- Finance and expense tracking