PCAM flashcards
PCAM · Build recall with a focused review session.
37 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Well-Architected Framework pillars
- Six pillars: Operational Excellence, Security, Reliability, Performance Efficiency, Cost Optimization, and Sustainability. Each pillar has design principles and best-practice questions scored by risk.
- AWS CAF perspectives
- Six perspectives: Business, People, Governance (business domain) and Platform, Security, Operations (technical domain). Used to structure cloud transformation planning and stakeholder responsibilities.
- Service control policy (SCP)
- An Organizations policy that sets the maximum permissions for member accounts. It filters permissions but never grants them; an explicit deny in an SCP overrides any allow.
- Landing zone
- A pre-configured, multi-account baseline with identity, logging, networking, and guardrails. AWS Control Tower automates its creation using Organizations, IAM Identity Center, and CloudTrail.
- The 7 Rs
- Migration strategies: Retire, Retain, Rehost, Relocate, Replatform, Repurchase, and Refactor/Re-architect. Choice depends on business value, effort, and application characteristics.
- Subnet AZ mapping
- A subnet belongs to exactly one Availability Zone and cannot span AZs. A VPC spans a Region and contains subnets in multiple AZs for fault isolation.
- Reserved IPs per subnet
- AWS reserves five addresses in every subnet: network address, VPC router, DNS, future use, and broadcast. A /24 therefore provides 251 usable addresses.
- Security group vs NACL
- Security groups are stateful, allow-only, and attached to ENIs. Network ACLs are stateless, support allow and deny, are evaluated in numbered order, and apply at subnet level.
- NAT Gateway behavior
- Provides outbound IPv4 internet access for private subnets, sits in a public subnet with an Elastic IP, and blocks unsolicited inbound traffic. It is AZ-scoped, so deploy one per AZ for resilience.
- Gateway vs interface endpoint
- Gateway endpoints (S3, DynamoDB) add a route-table entry at no hourly charge. Interface endpoints use PrivateLink ENIs and incur hourly and data processing charges.
- Transit Gateway
- A regional network hub connecting VPCs, VPNs, and Direct Connect gateways. It supports transitive routing, unlike VPC peering, and can peer across Regions.
- IAM policy evaluation logic
- Explicit deny overrides everything. Otherwise an allow must come from an SCP, identity or resource policy, and permissions boundary. Missing an allow means implicit deny.
- Permissions boundary
- A managed policy that sets the maximum permissions an identity-based policy can grant. It does not grant permissions and is commonly used for delegated administration.
- Cross-account role assumption
- The role's trust policy must allow the external principal, and the caller's identity policy must allow sts:AssumeRole. The caller then receives temporary STS credentials.
- KMS key rotation
- Customer managed symmetric keys support automatic annual rotation. Asymmetric keys and imported key material cannot rotate automatically and require manual rotation.
- S3 encryption options
- SSE-S3 uses AWS managed keys by default. SSE-KMS adds CloudTrail audit and per-key access control. SSE-C means the customer supplies and manages the key material.
- Compliance evidence sources
- AWS Artifact supplies compliance reports and agreements. AWS Config evaluates resource compliance, CloudTrail records API activity, and Security Hub aggregates findings across accounts.
- S3 durability and availability
- S3 Standard is designed for 99.999999999% durability and a 99.9% availability SLA. Durability reflects expected annual object loss; availability reflects successful request time.
- S3 storage class selection
- Intelligent-Tiering suits unknown or changing access. Standard-IA suits infrequent but rapid access. Glacier Instant Retrieval keeps millisecond access; Deep Archive is for long-term retention.
- EBS vs instance store
- EBS volumes are network-attached, AZ-scoped, and persist independently of instance lifecycle. Instance store is physically attached, ephemeral, and lost on stop or termination.
- DynamoDB partition limits
- Each partition supports up to 3,000 read capacity units and 1,000 write capacity units. A hot partition caps throughput regardless of the table's total provisioned capacity.
- RDS Multi-AZ vs replicas
- Multi-AZ keeps a synchronous standby for failover and does not serve reads. Read replicas use asynchronous replication, can serve read traffic, and can be promoted.
- Application Migration Service
- AWS MGN performs lift-and-shift rehosting by continuously replicating source servers at block level into AWS and launching them as EC2 instances with minimal cutover downtime.
- DMS and SCT
- Database Migration Service moves data with minimal downtime. Schema Conversion Tool converts schemas and code for heterogeneous engines, such as Oracle to Aurora PostgreSQL.
- Strangler fig pattern
- Incrementally replace legacy functionality with new services behind a facade, routing traffic gradually. Old and new coexist, reducing migration risk and enabling rollback.
- Containers vs serverless fit
- Lambda suits short-lived, event-driven, spiky workloads. Containers on ECS, EKS, or Fargate suit long-running or stateful processes needing runtime control and predictable concurrency.
- RTO and RPO
- Recovery Time Objective is the maximum acceptable downtime after a failure. Recovery Point Objective is the maximum acceptable data loss measured in time.
- DR strategy ordering
- From lowest cost and slowest recovery to highest cost and fastest: Backup and Restore, Pilot Light, Warm Standby, Multi-Site Active/Active. Each step improves RTO and RPO.
- Pilot Light vs Warm Standby
- Pilot Light replicates core data and keeps minimal services running, scaling up at failover. Warm Standby runs a scaled-down but fully functional copy of the workload.
- Static stability principle
- Failover relies on pre-provisioned capacity and does not depend on control-plane actions during an event. This avoids delays and dependency failures at the worst moment.
- Savings Plans scope
- Compute Savings Plans apply to EC2, Fargate, and Lambda across families, sizes, Regions, and OS. EC2 Instance Savings Plans are tied to a family and Region for deeper discounts.
- gp3 volume baseline
- gp3 provides a baseline of 3,000 IOPS and 125 MiB/s throughput, with IOPS and throughput configurable independently of volume size up to published maximums.
- S3 Intelligent-Tiering mechanics
- Automatically moves objects between access tiers based on observed usage. It charges a per-object monitoring fee but no retrieval fees, suiting unpredictable access patterns.
- CloudTrail vs Config vs CloudWatch
- CloudTrail records API activity for audit. AWS Config records resource configuration and changes for compliance. CloudWatch holds metrics, logs, and alarms for operations.
- Session Manager benefit
- Provides shell access to instances through Systems Manager without opening inbound ports, running bastion hosts, or managing SSH keys. Sessions are logged for audit.
- CloudFormation StackSets
- Deploys a single template as stacks across multiple accounts and Regions from one operation, commonly used to roll out governance and baseline resources organization-wide.
- Cost allocation tags
- Tags activated as cost allocation tags appear in billing reports, enabling cost breakdown by team, project, or environment. Tag policies in Organizations enforce consistent keys.