Skip to content

Modernize Infrastructure and Applications with Google Cloud

Course Overview

Technology moves fast. Modern organizations are shifting from rigid, local on-premises hardware to the elastic cloud compute model to gain speed, flexibility, reliability, and security. This course is part of the Google Cloud Digital Leader learning path.

Fundamentals of Modernization

To successfully modernize infrastructure and applications in the cloud, it is essential to understand core cloud computing terms and operational concepts.

Key Terminology

Compute : In the cloud, compute describes the ability of a machine to process information, handling tasks such as storage, retrieval, and analysis. Unlike traditional local servers, cloud computing uses a network of remote servers to offer on-demand access to resources, maximizing flexibility and scalability.

Workload : A specific application, service, or capability that can run in the cloud or on-premises (such as virtual machines, databases, or containers).

Abstraction : The architectural concept where a cloud provider handles the nitty-gritty details of hardware, networking, and software maintenance, leaving the customer to focus on their workloads.

Managed Services : Offerings where the cloud provider takes on the responsibility of managing and maintaining specific services (like databases, messaging queues, or analytics platforms), allowing organizations to focus strictly on business logic rather than routine operational maintenance.

Unlocking the Value of the Cloud

Shifting compute workloads from traditional on-premises environments to elastic cloud compute models offers several business and economic benefits:

  • Total Cost of Ownership (TCO): Reduces TCO by eliminating the need to purchase, house, and maintain physical IT infrastructure. Organizations shift from capital expenses (CapEx) with fixed capacity to a pay-as-you-go model (OpEx) where they pay only for what they use. Providers also offer committed use discounts and reservations for further savings.
  • Scalability: The ability to instantly adjust (increase or decrease) resources to meet the exact demands of an application (e.g., scaling up for spikes and scaling down to save costs).
  • Flexibility: Easily selecting, modifying, and adapting cloud services to best meet the organization's evolving business needs.
  • Reliability: High levels of uptime and availability guaranteed by global networks of redundant data centers. If one data center experiences a failure, others continue to operate.
  • Security: Multi-layered security features provided by the cloud provider, including encryption, advanced identity management, real-time threat detection, and assistance with regulatory compliance.
Podcast Transcript: Cloud Corner - Arjun's Case (Hesitant about moving workloads)

Joe: Welcome back to the Cloud Corner! We’re opening up the mailbag today, because we received a really excellent question from a listener named Arjun. Arjun is an IT director, and he writes: "My executive team is hesitant about moving our compute workloads to the cloud. They feel like we'd just be renting someone else's computers. How can I explain the actual business value to them beyond just 'renting servers'?"

Katelyn: Ya that’s a really common objection, Arjun. It’s easy to think of the cloud as just a hard drive in the sky, but the benefits go so much deeper.

Joe: They really do. So, Katelyn, if we’re building a case for Arjun, let’s start with the bottom line: Total cost of ownership or TCO.

Katelyn: Right. TCO isn't just the sticker price. It measures the cost of a system over its entire lifetime: purchase, maintenance, operations. When you move to the cloud, you eliminate the need to buy and maintain physical infrastructure. It moves you to a pay-as-you-go model. You only pay for the resources you actually use. Plus, cloud providers often offer discounts for long-term commitments, which can drive that TCO down even further.

Joe: The second big point for Arjun’s team is scalability. In the old days, if you expected a spike in traffic, you had to buy servers that sat empty half the year. But in the cloud, you can scale up instantly to meet demand and, just as importantly, scale back down when things are quiet to save money. That flexibility is huge.

Katelyn: Then, you have reliability. Cloud providers like Google have multiple data centers all over the world. If one goes down, others continue to operate. It gives you a level of uptime that is very hard to build on your own.

Joe: And we can’t forget security. It’s often a concern for leadership, but cloud providers offer massive security features: encryption, identity management, and real-time threat detection. This also helps heavily with regulatory compliance.

Katelyn: Now, Joe, there’s one more concept that I think is the "secret sauce" of the cloud: Abstraction.

Joe: Yes! Abstraction sounds technical, but it essentially means the cloud provider handles the nitty-gritty details of hardware, networking, and software maintenance.

Katelyn: And a specific part of that "secret sauce" is managed services. With managed services, the cloud provider takes on the responsibility of managing and maintaining specific services, such as databases, messaging queues, or analytics platforms.

Joe: That is a massive benefit. It allows the organization to focus on its core business logic and application development rather than getting bogged down in routine maintenance and operations.

Katelyn: Exactly. It frees your team up. Instead of fixing servers, they can focus on innovation. It helps you get products to market faster and gives you access to the latest tools the second they are released.

Joe: So, Arjun, tell your team it’s not just renting servers. It’s about agility, security, and freeing your team to build the future.

Katelyn: Thanks for the great question.

Cloud Migration Strategies (The 6 R's)

Before moving workloads, organizations begin with a Discovery and Assessment phase, inventorying the existing on-premises environment, analyzing applications, dependencies, and costs, to choose the right migration strategy[^1].

graph TD
    A[Discovery & Assessment] --> B[Retire]
    A --> C[Retain]
    A --> D[Rehost / Lift & Shift]
    A --> E[Replatform / Move & Improve]
    A --> F[Refactor / Re-architect]
    A --> G[Reimagine]

The 6 R's Defined

Retire : Removing unnecessary, non-cost-effective, insecure, or incompatible workloads from the platform.

Retain : Intentionally keeping workloads on-premises or in a hybrid cloud environment, continuing to manage them internally without cloud provider control.

Rehost (Lift and Shift) : Migrating a workload to the cloud without changing its code or architecture.

  • Pros: Simplest and fastest migration path.
  • Cons: Does not leverage cloud-native features; scaling and managing workloads unmodified can remain difficult.

Replatform (Move and Improve) : Migrating a workload while making minor changes to its code or architecture to leverage cloud-native scalability and reliability (e.g., migrating a database to a managed database service).

  • Pros: Increases workload performance and reduces operational expenses.
  • Cons: Modifying and testing workload code/architecture can be complex and time-consuming.

Refactor : Changing the code of a workload, such as transitioning it to a cloud-based microservices or serverless architecture.

  • Pros: Maximizes workload efficiency, scalability, and security to utilize full cloud capabilities.
  • Cons: Highly complex, expensive, and time-consuming.

Reimagine : Rethinking how an organization uses technology (like artificial intelligence and machine learning) to achieve its business goals, improving efficiency, agility, and customer experience.

Compute Architectures: Virtualization & Containers

Choosing the right compute foundation is critical for operational efficiency, scaling speed, and developer velocity.

Virtual Machines (VMs)

Virtualization is a resource optimization model allowing multiple operating systems to run concurrently on the same physical hardware, sharing a pool of processing, storage, and networking resources.

  • Compute Engine: Google Cloud's Infrastructure as a Service (IaaS) offering that allows users to run VMs on Google's infrastructure.
  • Pricing & Savings:
  • Sustained-Use Discounts: Applied automatically for VMs running more than 25% of a month.
  • Committed-Use Discounts (CUDs): Upfront commitments for 1 or 3 years in exchange for steep discounts.
  • Spot VMs: Utilization of spare Google compute capacity at discounts of up to 91%. Spot VMs can be preemptively reclaimed by Google at any time with a 30-second notice; they are ideal for fault-tolerant batch processing.
  • Custom Machine Types: Ability to tailor CPU, memory, and storage ratios precisely to avoid overallocating resources.

Containers

Containers package an executable application together with all its runtime dependencies (libraries, system tools, settings).

  • Comparison with VMs:
  • VMs: Virtualize the entire physical hardware stack via a Hypervisor, booting a full guest Operating System (OS). This is slower and heavier.
  • Containers: Virtualize the software layers above the host OS kernel. They share the host kernel, starting in milliseconds with a fraction of the memory footprint.
  • Portability: Containers run identically across local workstations, staging servers, and various cloud platforms.

Container Management & Orchestration

At enterprise scale, managing thousands of containers requires automated orchestration:

  • Kubernetes: An open-source container orchestration platform originally designed by Google for automated deployment, scaling, and recovery.
  • Google Kubernetes Engine (GKE): A fully managed Kubernetes service.
  • GKE Standard: Customer manages cluster nodes (Compute Engine instances) and configuration.
  • GKE Autopilot: Google manages the underlying nodes, cluster sizing, and security hardening automatically. Users are billed strictly per running Pod.
  • Cloud Run: A fully managed, serverless container platform. It abstracts away clusters entirely, scaling stateless containers automatically from 0 to thousands based on traffic. Ideal for lightweight services and web apps.

Serverless Computing

Serverless computing abstracts infrastructure entirely from the developer. The provider handles all provisioning, scheduling, scaling, and patch management.

Key Characteristics

  • Zero Management: Developers upload code or containers; they do not interact with operating systems or cluster configurations.
  • Pay-per-use Pricing: Zero cost for idle time. Billing is calculated strictly on CPU/Memory consumption during code execution.
  • Scale-to-Zero: Instantly scales down to 0 instances when traffic ceases.

Google Cloud Serverless Suite

  • Cloud Run: Runs containerized applications written in any language or library.
  • Cloud Run Functions: An event-driven platform for running single-purpose, lightweight snippets of code in response to cloud events (e.g., file uploads, database changes).
  • App Engine: A Platform as a Service (PaaS) to build and deploy web applications with integrated routing, versioning, and scaling.

Hybrid & Multicloud Management

Enterprises often retain legacy workloads on-premises or distribute applications across multiple cloud vendors to avoid vendor lock-in.

Key Definitions

  • Hybrid Cloud: An environment combining on-premises/private cloud infrastructure with public cloud services.
  • Multicloud: An environment utilizing services from multiple public cloud providers (e.g., Google Cloud, AWS, Azure).

Google Distributed Cloud (GDC)

GDC is a portfolio of hardware and software solutions that extends Google Cloud infrastructure and services (like GKE) to on-premises data centers and edge locations, managed through a unified control plane.

Data Management across Distributed Systems

  • Cloud SQL: Fully managed relational databases (MySQL, PostgreSQL, SQL Server) that simplify cross-environment replication and backup.
  • AlloyDB Omni: A downloadable, containerized edition of AlloyDB for PostgreSQL designed to run high-performance, AI-ready database clusters on-premises, at the edge, or in third-party clouds.
  • BigQuery Omni: A multicloud analytics engine that allows users to run BigQuery queries directly on data stored in Amazon S3 or Azure Blob Storage without moving or copying the data.
  • Looker: A cloud-agnostic Business Intelligence (BI) platform providing a single semantic layer for consistent metric definition across all data sources.

Application Programming Interfaces (APIs) & Management

An API (Application Programming Interface) is a structured set of specifications and protocols that allows different software applications to communicate with each other.

Apigee API Management

Apigee is an enterprise API management platform that acts as a secure proxy gateway between frontend consumers and backend services.

  • Key Features:
  • Security: Enforces API keys, OAuth 2.0 authentication, authorization, and encryption.
  • Traffic Management: Restricts traffic spikes and protects backend services using Throttling and Rate Limiting.
  • Analytics: Tracks developer usage, latency metrics, and error rates to optimize operations.
  • Developer Portals: Enables self-service registration and developer onboarding, turning internal APIs into monetized, tier-based products.

References