Skip to content

Solutions Architect – Professional · SAP-C02

SAP-C02 Tie-Breakers

The exam rarely asks what a service does. It asks which of four workable answers fits a constraint. This is a list of the distinctions that decide those questions, organised by the choice you are actually making.

How to read a question

Read the last sentence first. It carries the qualifier: MOST cost-effectively, LEAST operational overhead, MOST secure, with the LEAST downtime. That word decides between answers that are all technically correct.

Then find the constraint. It is usually one clause buried mid-scenario: a number (RTO, RPO, bandwidth, retention), a prohibition (cannot modify the application, no source code, vendor will not support changes), or an organizational fact (300 accounts, a central team, an auditor).

Two habits that pay:

Connectivity

NeedAnswer
Many VPCs plus on-premises, transitive routingTransit Gateway
Two VPCs, simple, low cost, no transitivity neededVPC peering
Expose one service, or overlapping CIDRsPrivateLink with an NLB endpoint service
Share subnets so teams deploy into a network you ownVPC sharing via RAM
Consistent bandwidth and latency to on-premisesDirect Connect
Cheap or fast-to-provision hybrid link, or DX backupSite-to-Site VPN
On-premises resolves AWS private zonesRoute 53 Resolver inbound endpoint
AWS resolves on-premises domainsRoute 53 Resolver outbound endpoint plus forwarding rules

Direct Connect resiliency, in the exam’s own language:

Transit Gateway is not transitive for free. Route table association controls which table an attachment uses for lookups. Propagation controls which routes get advertised into a table. Segmentation questions are answered by using more than one route table.

Failover mechanism

SignalAnswer
HTTP or HTTPS, caching, header manipulation at the edgeCloudFront with an origin group
TCP or UDP, static IPs required, clients ignore DNS TTLGlobal Accelerator
DNS-level routing by latency, geography, or weightRoute 53 routing policies
Sub-minute failover with no DNS dependencyGlobal Accelerator

If a question mentions clients caching DNS or hardcoding IPs, Route 53 is the wrong answer.

Disaster recovery

StrategyRPORTOWhat is running
Backup and restoreHoursHours (up to 24)Nothing
Pilot lightMinutesTens of minutesData replicating, compute off
Warm standbySecondsMinutesScaled-down full stack running
Multi-site active-activeNear zeroNear zeroFull stack in both, serving traffic

Map the numbers in the question to the row, then check cost. If the question says cost matters, the answer is the cheapest strategy that still meets the numbers, never the most resilient one.

Data replication choices:

Governance controls

ControlScopeGrants access?
SCPAccounts in an OU or the org, including the root userNo, only limits
Resource control policyResources in accounts, on the resource sideNo, only limits
IAM policyA principal in one accountYes
Permissions boundaryMaximum for one principal, set by an adminNo, only limits
Resource policyWho may touch this resourceYes, including cross-account
Session policyOne STS sessionNo, only limits

Rules the exam leans on:

Detection and security services

ServiceWhat it actually answers
GuardDutyThreat activity from CloudTrail, DNS, VPC Flow Logs, EKS and S3 telemetry
InspectorKnown CVEs in EC2 packages, ECR images, and Lambda functions
MacieSensitive data discovery and classification in S3
DetectiveInvestigating the scope of a finding across time and entities
IAM Access AnalyzerPolicies granting external access, and unused permissions
ConfigResource configuration compliance and remediation
Security HubAggregation and standards scoring across accounts and Regions
CloudTrailWho called which API and when
Audit ManagerEvidence collection mapped to a compliance framework

If the question is about a policy, it is Access Analyzer. If it is about behaviour, it is GuardDuty. If it is about packages, it is Inspector. If it is about data content, it is Macie.

Secrets and keys

Migration tooling

The 7Rs, with a one-line trigger each:

StrategyTrigger
RehostLift and shift, tight deadline, no changes
ReplatformSmall optimisation on the way, usually database to RDS
RepurchaseMove to SaaS, often no source code or vendor end of life
RefactorBusiness needs features the current architecture cannot deliver
RetireNobody uses it
RetainNot ready, or a dependency blocks it
RelocateVMware Cloud on AWS, move hypervisor-level with no conversion
JobTool
Inventory and dependency mappingApplication Discovery Service, tracked in Migration Hub
Server migrationApplication Migration Service
Ongoing DR for serversElastic Disaster Recovery
Database, same engineDMS
Database, different engineSCT for schema, then DMS with change data capture
Bulk one-time data over the wireDataSync
Bulk one-time data too large for the wireSnow Family
Ongoing scheduled file syncDataSync
On-premises apps needing cloud storage as a local shareStorage Gateway
Partners uploading over SFTP or FTPSTransfer Family
Faster uploads to S3 over long distancesS3 Transfer Acceleration

Bandwidth arithmetic, which appears at least once: at 1 Gbps fully utilised you move about 10 TB per day. Multiply by the utilisation the question allows, divide the data volume, compare to the deadline. If it does not fit, the answer is Snow.

Storage

NeedAnswer
Unknown or changing access patternsS3 Intelligent-Tiering
Known cold after N days, millisecond retrieval still requiredLifecycle to Standard-IA, then Glacier Instant Retrieval
Archive, minutes to hours acceptableGlacier Flexible Retrieval
Deep archive, 12 hours acceptable, lowest costGlacier Deep Archive
Immutable retention for complianceS3 Object Lock compliance mode, or AWS Backup Vault Lock compliance mode
Windows shares, NTFS ACLs, DFS, ADFSx for Windows File Server
Linux NFS shared across instancesEFS
HPC parallel throughput, S3-linkedFSx for Lustre
NetApp features, snapshots, multi-protocolFSx for NetApp ONTAP
Above 16,000 IOPS on one volumeio2 Block Express
General purpose with independently provisioned IOPS up to 16,000gp3

Vault Lock and Object Lock both have governance mode, which a sufficiently privileged principal can override, and compliance mode, which nobody can. If the question says “including administrators” or names an auditor, it is compliance mode.

Databases

SignalAnswer
Relational, multi-Region writesNot Aurora. Reconsider the requirement or use DynamoDB global tables
Relational, fast failover and read scalingAurora with reader endpoint and failover tiers
Key-value at any scale, single-digit millisecondDynamoDB
DynamoDB read hot spotsDAX
DynamoDB write hot spotsFix the partition key, add write sharding
Time seriesTimestream
Graph and relationshipsNeptune
Ledger with cryptographic verificationQLDB
Wide-column Cassandra workloadKeyspaces
MongoDB workloadDocumentDB
SQL Server licence cost is the problemAurora PostgreSQL with Babelfish
Analytics warehouse, repeated BI queriesRedshift
Ad hoc SQL over S3, infrequentAthena over partitioned Parquet

Connection storms from Lambda or a large fleet mean RDS Proxy. A repeated identical read query means ElastiCache.

Compute and purchasing

Application integration

NeedAnswer
Buffer work, one consumer group, retries and DLQSQS
Fan out one message to many subscribersSNS
Route events by content to many targets, cross-accountEventBridge
Strict ordering and exactly-once processingSQS FIFO
Multi-step workflow with state, retries, and human approvalStep Functions
Existing Kafka clientsMSK
Existing RabbitMQ or ActiveMQ clientsAmazon MQ
Replay of a stream by multiple independent consumersKinesis Data Streams

Cost levers, in rough order of impact

  1. Turn off non-production outside business hours.
  2. Rightsize using Compute Optimizer.
  3. Cover the steady baseline with Savings Plans.
  4. Gateway VPC endpoints for S3 and DynamoDB, which are free and remove NAT gateway processing charges.
  5. S3 lifecycle or Intelligent-Tiering.
  6. Reduce cross-AZ chatter between services.
  7. Spot for anything interruption-tolerant.

Gateway endpoints, for S3 and DynamoDB only, have no hourly or data charge. Interface endpoints have both, but still usually beat NAT gateway processing at volume.

Phrases that give the answer away

Phrase in the questionWhere to look
Least operational overheadThe most managed service on the list
Even administrators must notSCP, or compliance mode lock
Cannot modify the applicationEFS, RDS Proxy, ALB, a sidecar, an ASG of one
No source code availableRehost or repurchase, never refactor
Overlapping CIDR blocksPrivateLink
Hundreds of accountsOrganizations, Control Tower, delegated administrator, StackSets, Firewall Manager
Existing objects or existing dataBatch operations or a backfill job, because live replication is forward-only
Clients cache DNSGlobal Accelerator
Sub-second replication, writes in every RegionDynamoDB global tables
Unpredictable access patternsIntelligent-Tiering, or on-demand capacity
Must prove the key never left our controlKMS custom key store on CloudHSM
Within the next 30 days, large volumeDo the bandwidth arithmetic, expect Snow

Independent study material written from publicly available sources. Not affiliated with or endorsed by Amazon Web Services.