AWS Solutions Architect Professional (SAP-C02) - Full Practice Exam 1

1. A business-critical relational application needs automatic recovery from an AZ failure. Read scaling is not currently required. Which baseline design best meets this requirement?
2. A business must restrict resource creation to approved Regions but still support required global services. What is the best implementation approach?
3. A business unit needs a private partner TCP service. The partner and customer use overlapping CIDRs, and neither wants broad network connectivity. Which design is best?
4. A central network account owns a Transit Gateway used by production and development VPCs. Both environments need a shared DNS VPC, but must not route to each other. Which design is appropriate?
5. A central team must deploy a common IAM and logging baseline into new accounts in selected OUs, including accounts created later. Which tool most directly supports this deployment model?
6. A company cannot attribute growing cloud cost to product teams because resource tagging is inconsistent. Which improvement creates a more useful financial operating model?
7. A company has 120 AWS accounts. Product teams own deployments, but security must prevent disabling organization audit logging. Teams still need different workload permissions. Which design best separates the organizational boundary from workload grants?
8. A company has daily database backups but has never restored them. Management asks whether the system meets a two-hour RTO. What is the most useful next step?
9. A company has finished rehosting and wants to prioritize modernization. Which method best identifies valuable next steps rather than rewriting everything?
10. A company has hundreds of independently maintained copies of common infrastructure. It needs versioned reusable products while preserving team deployment autonomy. Which service most directly supports a governed internal catalog?
11. A company has many dependent applications and no reliable inventory of server relationships. It must group migrations to reduce broken dependencies. What should it do first?
12. A company hosts a private S3 origin behind CloudFront. It must prevent direct public object reads while allowing only its distribution to retrieve content. Which configuration is preferred?
13. A company migrates Oracle to a different supported database engine. It needs to identify schema features that require conversion before planning data transfer. Which capability is most relevant?
14. A company migrates to a new account and Region. An application works in testing but fails after production cutover due to a low service quota. How should future waves improve?
15. A company must move 300 virtual machines to AWS with minimal application changes under a short data-center exit deadline. Which migration approach should it evaluate first?
16. A company must replicate object data to a recovery Region. Its RPO cannot be assumed zero, and administrators must know whether the replication goal is met. Which design is appropriate?
17. A company must survive loss of a Direct Connect facility without losing hybrid access. Which architecture is most appropriate?
18. A company needs a graph query that explores several levels of relationships between accounts, devices, and transactions with variable traversal depth. Which managed database is the most direct fit?
19. A company repeatedly discovers overlapping CIDRs across independently created VPCs. Future acquisitions must be easier to integrate. Which governance improvement is most appropriate?
20. A company requires an application recovery time under one hour after regional loss but can tolerate several minutes of data loss. It cannot fund a fully active duplicate. Which strategy should be evaluated first?
21. A company runs a constant EC2 baseline all year and wants to reduce cost without changing workload architecture. Usage is well understood, but instance families may change. Which approach should be evaluated?
22. A company shares a KMS-encrypted S3 dataset from account A with an analytics role in account B. The bucket policy permits the role, but reads fail at decryption. What additional authorization is required?
23. A company wants an enterprise access model based on project attributes across many roles and resources. Which prerequisite is essential for trustworthy ABAC?
24. A compute workload has a steady baseline and large interruptible batch peaks. Checkpoints allow interrupted tasks to resume. Which purchasing strategy best balances cost and reliability?
25. A container platform runs a different application image under the same mutable tag in staging and production. How should release reproducibility be improved?
26. A customer portal stores user sessions on individual web servers. Instances scale across AZs and are frequently replaced. Which design best supports stateless application scaling?
27. A database migration must keep the source running during most of the transfer and minimize the final outage. Which pattern is appropriate for a supported source and target?
28. A document platform serves immutable public images globally. The origin receives repeated requests for the same content and users report high cross-continent latency. What design most directly improves both origin load and delivery latency?
29. A file-processing application requires a shared POSIX file system mounted concurrently by Linux instances in several AZs. Which service most directly fits?
30. A financial processing API receives retried client requests after timeouts. An accepted request must create at most one order, even if the response is lost. Which design is best?
31. A global application accepts writes in multiple Regions and needs a DynamoDB replication design. The application cannot ignore cross-Region consistency and conflict behavior. What should the architect do?
32. A global company wants centrally governed branch and multi-Region VPC connectivity using network segments and policy-driven topology. Which managed service should it evaluate first?
33. A global enterprise is distributing shared software artifacts to accounts in multiple Regions. Releases must remain traceable to tested immutable versions. Which approach is most appropriate?
34. A global order service needs low-latency reads near users, but all writes must preserve a single relational transaction boundary. Regional recovery must have a small data-loss window. Which database design best fits these stated constraints?
35. A global UDP application needs static front-door IPs and health-based traffic steering to two Regions. HTTP caching is not required. Which architecture is best?
36. A hybrid enterprise needs VPC clients to resolve on-premises names and on-premises clients to resolve private AWS names. Which two components address the DNS directions? Select 2 answers.
37. A legacy batch processor is being moved to managed compute. Tasks have dependencies, varying resource needs, and can run for hours. Which service should be evaluated rather than forcing every task into a short Lambda execution?
38. A migration cutover plan has validated the target database but not application data reconciliation. Which two checks are most important before declaring success? Select 2 answers.
39. A migration team needs a reversible cutover from an old API to a new implementation. Both versions can serve the same contract, and traffic must increase gradually only while metrics remain healthy. Which design fits?
40. A modernization project separates order processing from customer notifications. Notification failure must not roll back accepted orders, and each service must scale independently. Which interaction is preferred?
41. A monolith must be modernized gradually while its existing APIs remain available. A full rewrite is too risky. Which pattern best supports incremental replacement?
42. A new analytics platform stores large append-only datasets and runs periodic SQL analysis. The team wants minimal database administration and low storage cost. Which design is appropriate?
43. A new asynchronous service processes variable-duration jobs. Workers may fail after a job starts, and work must be retried without losing it. Which two design elements are appropriate? Select 2 answers.
44. A new backend needs reliable publish/subscribe delivery to several independently scaled worker queues. Each consumer must be able to fall behind without blocking the others. Which design fits?
45. A new service needs per-item conditional updates at very high scale and predictable key-based reads. Relational joins are not required. Which database design is most appropriate?
46. A platform serves highly sensitive tenant-specific responses through CloudFront. A cache configuration accidentally ignores an authorization-dependent response variation. What should be done?
47. A platform team shares centrally owned VPC subnets with application accounts. What remains important to preserve responsibility boundaries?
48. A production ALB fleet loses requests during instance scale-in because applications need 45 seconds to finish active work. Which improvement best addresses graceful removal?
49. A recovery design copies a database but omits secrets, KMS permissions, quotas, and external dependencies in the recovery Region. What is the best assessment?
50. A recovery test succeeds only because a particular engineer manually changes undocumented settings. What improvement best makes the RTO repeatable?
51. A regulated company is introducing a multi-account landing zone. It requires isolated log and audit accounts, governed account provisioning, and controls applied consistently as accounts grow. Which approach minimizes custom governance plumbing?
52. A relational application repeatedly opens short database connections from Lambda and experiences connection pressure during bursts. Which improvement should be evaluated?
53. A retrying client worsens an overloaded downstream service by immediately sending all failed requests again. Which change best improves stability?
54. A security team requires organization-wide centralized logs, while workload teams must not be able to erase the retained evidence. Which two controls are appropriate? Select 2 answers.
55. A serverless API sees sudden bursts that overwhelm a relational database. Requests can be accepted asynchronously and processed within minutes. Which architecture best absorbs the burst?
56. A serverless application uses a standard queue. After a timeout, a message is processed twice and creates duplicate downstream records. Which change is most robust?
57. A service needs end-to-end tracing across an API, asynchronous queue processing, and downstream calls. Which design is appropriate?
58. A single large EC2 instance hosts both the web tier and a background batch process. Batch peaks degrade customer latency. Which improvement best reduces resource coupling?
59. A team is moving an NFS dataset to AWS while the source remains active. It wants managed incremental transfer over the available network with validation. Which service should it evaluate?
60. A team moves a Windows application that requires shared SMB storage and Windows access semantics. Which managed storage should be evaluated first?
61. A team optimizes average API latency, but customer complaints correlate with extreme slow requests. Which measurement should guide the next investigation?
62. A vendor assumes customer roles using its own AWS account. Customers need protection against another vendor customer causing access to their roles. Which trust design addresses this?
63. A web application reads the same expensive query result thousands of times per minute. Slightly stale results are acceptable for 30 seconds. Which optimization best reduces database load?
64. A workload writes large amounts of private S3 traffic through NAT. It runs in a Region where an S3 gateway endpoint is supported. Which change should be evaluated to reduce unnecessary NAT processing?
65. An application database has grown, and an unindexed filter now scans most rows for each request. Which improvement should precede an expensive instance upgrade?
66. An application handles bursts poorly because scaling follows average CPU, while the actual bottleneck is queued work per worker. What is the best improvement?
67. An application moving to containers currently embeds database credentials in its deployment archive. What should the modernization include?
68. An audit team needs centralized configuration visibility across accounts and Regions, but remediation must follow approved per-account workflows. Which design fits?
69. An enterprise uses its corporate identity provider for workforce access. Administrators need different permissions in 40 accounts and should not manage IAM user keys. Which design is preferred?
70. An existing cross-Region application has a read replica but recovery fails because clients continue writing to the old primary. What must be added to the recovery design?
71. An existing DynamoDB table throttles despite substantial unused table capacity. Most requests use one tenant ID as the partition key. What is the best improvement?
72. An image workflow contains several managed service steps, conditional branches, and recoverable failures. Operations needs per-execution history and explicit retries without a custom workflow engine. Which design fits?
73. An IoT ingestion pipeline has many devices and several independent consumers needing ordered processing within each device stream and the ability to replay retained events. Which service fits?
74. An S3 analytics workload scans terabytes of JSON to answer queries for one date and a few columns. Which two changes most directly reduce scanned bytes? Select 2 answers.
75. Two versions of an application run concurrently during migration. The new version renames a database field that the old version still requires. Which rollout pattern best preserves compatibility?

Comments — 0