Prepare and test your skills

Prepare and test your skills

Controlling access with Identity and Access Management (IAM) means every person and workload gets only the permissions needed for its job. Cloud Identity centrally manages users and groups and enables single sign-on to Google Cloud and other applications. The IAM model connects four parts where a policy attached to a resource grants a role to a principal.
When a request arrives, Google Cloud checks the allow policy on that resource. The request succeeds only if the policy binds that principal to a role containing the required permission. Least privilege shrinks harm because a compromised account can reach fewer resources.
Exam tip: A role defines what is allowed, while an IAM allow policy defines who gets that role and on which resource.
Organizing resources and enforcing guardrails means using hierarchy and automatic limits to separate workloads and control where they run. The Organization resource sits at the top and gives central visibility, and it contains folders that group related projects. Projects then contain service resources such as Compute Engine virtual machines, Pub/Sub topics, Cloud Storage (opens in a new tab) buckets, and Cloud VPN endpoints. Sensitive data with special access or residency needs stays in dedicated Assured Workloads folders or projects so stricter controls apply in one place.
Guardrails apply automatically across folders and projects. Organization policies set constraints such as resource location limits that keep workloads in approved regions. VPC Service Controls create a security perimeter around sensitive services where traffic inside the perimeter can flow between protected services while requests crossing the perimeter are denied unless explicitly allowed. Policy Intelligence tools show who has access so overly broad grants can be found and fixed.
An Euler-style set diagram showing a VPC Service Controls security perimeter containing protected projects and services, where traffic between services inside the perimeter flows freely, requests crossing the boundary are denied unless explicitly allowed, and internet clients and unapproved projects remain outside.
Monitoring operations and protecting sensitive data means recording activity, finding threats early, and keeping regulated data in protected places. Security Command Center is the centralized service for continuous monitoring, threat detection, and response. It combines threat intelligence, detection, and web scanning to find misconfigurations, vulnerabilities, and active threats. Because misconfiguration is a leading cause of breaches, continuous monitoring gives warning before a weak setting becomes an incident.
Findings flow from Google Cloud resources to Security Command Center and logging systems, then back to administrators who fix IAM policies, organization policies, or perimeters. Access Transparency logs show qualifying access by Google personnel, including support and outage recovery. Sensitive data such as account information, health information, personal identifiers, cardholder data, and ID numbers must stay inside compliant folders and must not be placed in resource names, project names, labels, descriptions, URIs, file paths, or firewall rules.
Modern threats and cloud defenses explain how attacks start, what they cost, and why identity-based controls replace fixed perimeters. Ransomware is sequenced malicious software that encrypts, steals, or blocks data until payment, often in cryptocurrency for a decryption key. The typical attack moves in order from reconnaissance to intrusion through phishing or software vulnerability, to installation and contact with command control, to spread and encryption with exfiltration, to ransom demand. Business impact includes downtime and response cost, legal penalties under rules such as PCI DSS, GDPR, CCPA, FedRAMP, and CJIS, and lasting loss of customer trust.
Traditional security trusts users inside a physical perimeter guarded by badges, cameras, fences, and edge hardware appliances. Google Cloud uses zero trust where no user or device is trusted by location, so every request is authenticated and authorized based on role, device posture, and location. Confidentiality keeps Customer Data private by default, integrity blocks unauthorized change with approvals such as Binary Authorization, and availability keeps services running through shared duties. Authentication proves who is requesting, authorization decides what is allowed through IAM roles, and Cloud Audit Logs record Admin Activity, Data Access, System Event, and Policy Denied events for review.
| Service model | Google Cloud manages | Customer manages |
|---|---|---|
| Infrastructure as a service (IaaS) | Hardware, storage, and network | Data, applications, operating systems, and access |
| Platform as a service (PaaS) | Platform and infrastructure including patching | Data, access, code, images, and IAM configuration |
| Software as a service (SaaS) | Everything except data and its access and usage | Data and its access and usage |
Continue with this study guide: Trust and security in the cloud
Defense in depth in Google Cloud stacks independent layers so failure of one control does not expose data. The layers span physical facilities and custom hardware, the global network, data encryption, identity and access, and security operations. Google secures the underlying infrastructure while the customer secures data and how data is accessed and used. Because encryption, network defense, and monitoring are built in by default, organizations need fewer add-on tools.
Physical protection starts at the data center perimeter with fencing, vehicle barriers, alarms, access cards, cameras, and guards. Entry to the data center floor requires a security corridor with badge and biometric checks open only to approved roles. Inside, custom servers omit unneeded parts, vendors are vetted, Titan security chips verify boot code, and redundant power, cooling, and fire protection support continuous operation.
The private global network is Google’s worldwide network built from Google-owned fiber, public fiber, and undersea cables. The network connects directly to most internet providers, so customer traffic enters Google’s protected network early instead of traveling far over the public internet. Each avoided hop across providers removes another chance for interception. Traffic flows from customer devices to Google Front End (GFE) servers at the edge, where harmful requests and distributed denial-of-service (DDoS) attacks are filtered before reaching production systems.
Google data centers are custom facilities designed for power usage effectiveness (PUE), which measures how efficiently electricity is used. Every critical power component has primary and alternate sources, backup generators can run a facility at full capacity, and cooling and fire systems maintain safe operation. Custom servers omit unneeded video cards and connectors that waste power and can create risk, storage uses full disk encryption and drive locking, and retired drives are verified as sanitized or shredded and recycled.
Exam tip: Connections between virtual machines that use internal IP addresses are automatically encrypted, while connections that use external IP addresses leave Google’s network and are not automatically encrypted.
Encryption in Google Cloud protects data in all three states by making content unreadable without correct keys. Data at rest is encrypted by default at the application layer, storage device layer, or both, using keys held in Keystore that rotate automatically. Data in transit is protected with Transport Layer Security (TLS) and internal controls, while data in use requires isolated processing because it lives in memory during computation.
| Option | Who manages keys | Customer control |
|---|---|---|
| Google default encryption | Google in Keystore | Automatic protection with no key visibility |
| Customer-managed encryption keys (CMEKs) | Customer in Cloud KMS | Location, access, rotation, use, and destruction |
| Customer-supplied encryption keys (CSEKs) | Customer provides material per request | Direct control of key material held in memory |
| Cloud External Key Manager (Cloud EKM) | Customer in third-party system | Separation of data and keys outside Google |
Traffic flows from users on the internet to GFE, which terminates HTTP(S), TCP, and TLS connections worldwide and routes traffic over Google’s network. Inside Google, VM-to-VM traffic within a Virtual Private Cloud (VPC) (opens in a new tab) network or across peered VPCs is encrypted at the network layer, while Application Layer Transport Security (ALTS) provides mutual authentication and encryption between services. Confidential Computing protects data in use by running work inside a Trusted Execution Environment with hardware isolation and attestation, including Confidential VM and Confidential GKE Nodes.
Authentication, authorization, and auditing form a connected chain that proves identity, grants permitted actions, and records what happened. Corporate identities are centrally managed with Cloud Identity and Google Workspace, while service accounts represent workloads such as code on a virtual machine. People sign in with managed credentials after running gcloud auth login, strengthened by strong passwords, multi-factor authentication (MFA) for all users, and federation with providers such as Active Directory.
Authorization is enforced by IAM where a permission allows one operation and a role collects permissions granted to a principal on a resource. Permissions cannot be granted directly, and a child resource inherits allow policies from its parents in the organization to folder to project chain. An IAM deny policy overrides any allow policy on the same resource and descendants, except deny policies cannot be set on Cloud Storage buckets or objects. Two-Step Verification (2SV) adds a second proof after the password, such as a physical security key tap or mobile prompt approval, so a stolen password alone does not grant access. Cloud Audit Logs create an immutable record of configuration changes and blocked access that is viewed in Logs Explorer and exported with log sinks to BigQuery.
Network attack protection filters unwanted traffic at the edge before it reaches backends. A Google Cloud Armor security policy is a set of rules that allows or denies incoming traffic based on IP address, geography, and reputation. The policy acts at GFEs, so malicious requests never contact backends. Policies attach to load balancer backend services behind an external Application Load Balancer, and traffic flows from the client to GFE, through the policy check, then from the load balancer to the backend only if allowed.
Global capacity absorbs volumetric attacks where one source or many compromised hosts try to overwhelm a service. Cloud Load Balancing uses a single anycast IP to front instances in all regions and directs each user to the closest backend with capacity. Cloud Armor adds Web Application Firewall rules aligned with OWASP to block cross-site scripting, SQL injection, and file inclusion and code execution attempts. VPC firewall rules then enforce the final boundary by permitting only load balancer and health-check traffic to reach backend VMs. Google Security Operations completes the workflow with collection into a Unified Data Model, detection with threat intelligence, investigation with search and case management, and response with automated playbooks.
Continue with this study guide: Google’s trusted infrastructure
Google Cloud earns trust by denying personnel access by default, limiting any grant, and making access visible. Protection of customer data is a primary design consideration, customers own their data completely, and Google does not use customer data for advertising or other purposes beyond contracted services. Customers run on the same infrastructure as Search, YouTube, and Gmail, so they benefit from the same large-scale controls.
Transparency tools show what happened and prove controls work. Access Transparency records actions Google personnel take on Customer Data, including resource, action, time, reason, and accessor details. Cloud Audit Logs record actions by members of the customer’s own organization, so the two logs together cover both Google administrative access and customer actions. Access Approval lets customers approve or deny Google personnel requests before access to covered data, and approved requests link to Access Transparency logs. Key Access Justifications adds control for customer-managed keys by recording each external key use.
Access is allowed only for limited reasons after strict checks, including placement on access control lists, trusted device, MFA with Titan Security Key, justification such as a support ticket, and extra approval when required. Customer-initiated support is tied to a ticket and limited to low-sensitivity data, outage recovery allows logged access for backup without bypassing allowlists, and legal or safety reviews are rare and tightly controlled.
Exam tip: Access Transparency shows Google personnel access after it happens, while Access Approval controls that access before it happens.
The shared responsibility boundary divides duties so each side knows which controls it must provide. The general rule is that customers secure what they bring to the cloud while Google protects the cloud itself. The exact split changes by service type because Google manages more of the stack as services become more complete. Shared fate extends this split into ongoing partnership with secure landing zones, blueprints, and best-practice guidance.
| Service model | Google Cloud responsibility | Customer responsibility |
|---|---|---|
| IaaS | Hardware, storage, and network | Everything above, including data, applications, and access |
| PaaS | Runtime and platform layers plus infrastructure | Application controls, IAM management, data security, and client protection |
| SaaS | Everything except data and its access and usage | Data and its access and usage |
Customer duties center on data, access, and configuration. Customers must know which data has residency needs, place sensitive data in appropriate Assured Workloads folders, configure IAM so only appropriate personnel can access resources, and avoid placing sensitive data in names, IDs, labels, URIs, or firewall rules. Google duties center on infrastructure, enforcement, and oversight, including default encryption, physical and network security, supplier assessment, and enforcement of customer IAM and Assured Workloads controls. Policy enforcement flows from the Organization resource through folders and projects to service resources, where IAM allow policies bind principals to roles.
Data location controls help organizations meet legal and organizational rules about where data lives and which laws govern it. Data residency is the practice of keeping data within specific geographic boundaries, while data sovereignty is the legal and jurisdictional control over data. Residency answers where data is stored and processed, while sovereignty answers which country’s laws and authorities govern it. Residency controls help satisfy sovereignty obligations but do not replace them.
| Concept | Main question | Google Cloud example |
|---|---|---|
| Data residency | Where is data stored? | Regional secrets kept in one location at rest, in use, and in transit |
| Data sovereignty | Which laws govern data? | Local licensing and oversight rules for a regulated region |
Zones, regions, and multi-regions define how widely data spreads. A zonal resource such as a Compute Engine VM lives in one zone, a regional resource spreads redundantly across zones in one region for higher availability, and a multi-regional resource distributes across regions to survive loss of one region. Secret Manager enforces residency through regional secrets that stay within that location and are accessed only from the same location, unlike global secrets replicated for availability. Cloud Storage and BigQuery apply control through bucket or dataset location, where a regional choice keeps data inside one national boundary and a multi-regional choice replicates across regions for resilience.
Governance enforces approved geographies in order. First the administrator creates an Assured Workloads folder and selects a control package. Then the package sets the Resource Locations constraint so only approved regions can be chosen. Then workloads created inside the folder inherit those boundaries, and requests outside allowed regions are blocked.
Proving compliance means matching controls to laws and showing auditors that they work. Independent third-party audits verify Google controls because the auditor is independent from the provider. Key standards include ISO/IEC 27001 for information security management, ISO/IEC 27017 for cloud security, ISO/IEC 27018 for cloud privacy, ISO/IEC 27701 for privacy, SOC 2 and SOC 3 reports, FedRAMP for the United States government, BSI C5 for Germany, and MTCS for Singapore. Transparency reports also disclose government requests and service disruptions, and Google policy directs governments to request data from the customer and notifies the customer unless prohibited.
The Compliance Manager (opens in a new tab) framework support and resource center help teams discover which services support each regulation and find current proof. The Compliance resource center is the central repository for certifications, attestations, and offerings used first during assessment. Compliance Reports Manager is the library that supplies the actual downloadable audit reports for internal risk teams and external auditors. Audit readiness then moves in order from assessing obligations, to automating implementation with Assured Workloads and blueprints, to continual monitoring with Audit Manager, Security Command Center, Cloud Audit Logs, and Access Transparency validated by an external auditor.
Continue with this study guide: Google Cloud’s trust principles and compliance
A role is a collection of permissions that defines what is allowed, while an Identity and Access Management (IAM) allow policy defines who gets that role and on which resource. The allow policy binds a principal such as a Google Account, service account, or Google Group to the role on a specific resource, and a request succeeds only if the policy binds that principal to a role containing the required permission.
Access Transparency logs record actions Google personnel take on Customer Data, while Cloud Audit Logs record actions by members of the customer’s own organization. Together the two logs cover both Google administrative access and customer actions, with Access Transparency showing Google personnel access after it happens.
Data residency is the practice of keeping data within specific geographic boundaries and answers where data is stored and processed, while data sovereignty is the legal and jurisdictional control over data and answers which country’s laws and authorities govern it. Residency controls help satisfy sovereignty obligations but do not replace them.
Last updated