AWS Certified Solutions Architect - Associate SAA-C03 Practice Test 80
Question 790
An existing VPC's private subnet is almost full of allocated addresses. The team proposes changing its CIDR mask in place from /26 to /20 while preserving all existing subnet resources. Which constraint matters?
- Edit the subnet CIDR directly while hundreds of ENIs remain and assume AWS will transparently renumber them.
- Plan supported VPC CIDR expansion and a new appropriately sized subnet or migration; an occupied subnet cannot simply be resized in place.
- Reduce the number of Route 53 TXT records to increase the subnet's IPv4 address count.
- Increase EC2 vCPU quotas and assume each subnet gains more available IPv4 addresses.
Explanation
Correct Answer
B:AWS subnet CIDR blocks are fixed after creation; expansion generally requires added address space and replacement/redistribution.
Incorrect Answers
A:Subnet address ranges are not arbitrarily modifiable in place.
C:DNS records are unrelated to the subnet's allocated IP capacity.
D:Compute quota changes do not expand a subnet's CIDR range.
Question 791
A Fargate service repeatedly fails to create new tasks despite ample CPU quota. The private subnet has fewer available IPv4 addresses than the desired count of task ENIs and supporting endpoints. What resource should the architect plan?
- Increase only the ECS desired count while retaining the exhausted subnet CIDR.
- Reduce the number of S3 bucket lifecycle rules to generate more task IP addresses.
- Provide adequate subnet IP address capacity for task ENIs, load balancers and other managed interfaces, potentially using additional subnets.
- Raise EC2 CPU credit balance and assume Fargate no longer needs network interface addresses.
Explanation
Correct Answer
C:Many managed workloads consume ENIs/IPs; compute demand planning must include address capacity.
Incorrect Answers
A:Additional desired tasks cannot allocate addresses that do not exist.
B:Object lifecycle policy has no influence on subnet IPv4 allocation.
D:Fargate task networking still needs the configured VPC addresses.
Question 792
A dual-stack VPC gains IPv6 addresses for an internal service, but a legacy partner requires IPv4-only connections and the IPv4 subnet is exhausted. What should be acknowledged?
- Assume assigning an IPv6 CIDR automatically doubles the IPv4 address count in each subnet.
- Remove the IPv4 route table and expect IPv4-only clients to connect over unmodified IPv6 sockets.
- Enable DNS64 and NAT64 only and presume that an IPv4-only partner can initiate a connection to any IPv6-only server without additional compatible design.
- Plan adequate IPv4 capacity or supported translation/connectivity paths independently; adding IPv6 does not itself enlarge the IPv4 pool.
Explanation
Correct Answer
D:Dual-stack adds another address family but cannot make IPv4-only endpoints use IPv6 addresses without suitable translation.
Incorrect Answers
A:An IPv6 VPC CIDR assigns addresses in the independent IPv6 space and does not expand the IPv4 CIDR or available IPv4 subnet addresses. An IPv4-only integration still needs a supported IPv4 or translation path.
B:An IPv4 client cannot reach an IPv6-only destination without compatible network translation or dual-stack support.
C:DNS64/NAT64 generally support IPv6 clients reaching IPv4 services; they do not automatically give an IPv4-only partner unrestricted access to an IPv6-only service. IPv4 address planning or a suitable translation path is still required.
Question 793
A retail site runs edge-facing web origins near Paris and Milan. Marketing wants traffic generally routed by geographic closeness but deliberately wants Paris to serve a larger surrounding area during a seasonal campaign, with both endpoints healthy. It does not want the same fixed 20/80 fraction for every resolver. Which routing policy supplies the intended geographic bias?
- Use Route 53 geoproximity routing with a suitable positive bias for the Paris resource and verify how the geographic catchments change.
- Use Route 53 weighted routing because fixed percentages directly enlarge a geographic catchment around a specific location.
- Use failover routing, which sends every resolver to Paris until Paris is unhealthy and never selects by geography.
- Use a long DNS TTL to redefine geographic boundaries independently of resource locations.
Explanation
Correct Answer
A:Geoproximity routing considers resource and requester locations, with a bias adjustment used to expand or shrink the area routed to a resource. Weighted routing expresses answer proportions but not geographic catchment areas.
Incorrect Answers
B:A weighted record changes relative frequency of answers and cannot express a location-based catchment expansion.
C:Failover is active-passive by health and does not implement adjustable regional geographic preference.
D:TTL affects DNS caching duration, not the geoproximity boundaries selected by the routing algorithm.
Question 794
A legacy client can retry any of several backend IPs returned by one DNS lookup. The platform has six independently hosted endpoints, and the architect wants unhealthy endpoints omitted from future DNS responses without placing them behind an ELB. Which Route 53 record policy fits this request?
- Configure latency routing, which always returns a complete set of six healthy endpoint IPs in one answer.
- Configure Route 53 multivalue-answer routing with separate records and suitable health checks for the six endpoints.
- Configure one fixed simple A record with no health check and expect automatic exclusion of unhealthy addresses.
- Configure a two-record primary/secondary failover policy and expect it to return all six healthy endpoint addresses to clients simultaneously.
Explanation
Correct Answer
B:Multivalue-answer routing can answer with up to eight health-evaluated records, enabling clients to retry another returned IP. It improves DNS-level availability but is not a replacement for application load balancing.
Incorrect Answers
A:Latency routing is designed to select suitable lower-latency regional destinations, not to guarantee a set of all healthy IPs.
C:A fixed simple record without health evaluation does not automatically remove failed endpoints.
D:Failover routing is designed to prefer a primary and use a secondary on health failure. It is not equivalent to multivalue-answer routing, which can return several independently health-checked addresses in one response.
Question 795
A publishing service must serve country-specific legal content according to the requester's country, even when a farther Region has lower latency. Which Route 53 routing category directly matches this policy?
- Use latency-based routing solely because it always returns a server in the user's country.
- Select Route 53 geoproximity routing with resource bias and assume it guarantees the country-specific legal selection for every DNS resolver.
- Use Route 53 geolocation routing with records and default handling reflecting the required country mapping.
- Use simple round-robin routing and assume every returned Region applies the correct local legal constraints.
Explanation
Correct Answer
C:Geolocation policies choose DNS answers based on a defined origin geography, distinct from simply minimizing latency.
Incorrect Answers
A:Latency preference is not a legal geographic-routing guarantee.
B:Geoproximity bias steers traffic based on resource and resolver geography; its adjustable catchment boundaries do not enforce a precise legal country-to-content mapping in the way geolocation records can.
D:Without country-specific rules, users may be sent to prohibited content regions.
Question 796
A video provider knows the exact public CIDR ranges used by three contracted ISPs. It must route DNS queries from those network ranges to negotiated delivery endpoints even when a general continent geolocation lookup would choose differently. Queries from unmatched ranges need a fallback answer. Which Route 53 configuration satisfies that explicit mapping?
- Use Route 53 geolocation routing by country only, assuming every ISP CIDR remains within an accurate country catchment indefinitely.
- Use weighted routing with equal weights for all endpoints because the weight identifies the query source ISP CIDR automatically.
- Use health-based failover records alone and assume primary/secondary health checks classify which ISP originated each DNS query.
- Configure Route 53 IP-based routing records tied to a CIDR collection of ISP ranges and a default location response for unmatched IP ranges.
Explanation
Correct Answer
D:IP-based routing applies customer-defined CIDR-location mappings rather than relying only on AWS geographic estimates. A default location can handle queries that do not match an explicit CIDR rule; resolver/ECS behavior still needs testing.
Incorrect Answers
A:Geolocation relies on inferred client or resolver geography, which may not capture explicit ISP range-to-endpoint contracts or organizational overrides. Customer-managed CIDR collections provide the requested finer mapping.
B:Weighted records distribute DNS answers in defined proportions. They do not evaluate custom client-IP CIDR collections or ensure each contracted ISP reaches its designated endpoint.
C:Route 53 failover controls primary-versus-secondary health response, not contractual CIDR-based endpoint selection. Healthy records still require a distinct policy to map specified IP ranges.
Question 797
A data-ingestion team streams web events with well-distributed, high-cardinality partition keys. Traffic alternates between modest baseline and irregular multi-hour peaks, and the team wants AWS to manage shard capacity rather than operators predicting a fixed shard count. Which Kinesis capacity mode fits if the team still implements retries for sudden beyond-peak bursts?
- Use Kinesis Data Streams On-demand Standard mode, monitor peak-growth behavior and retry temporary throttling when demand outruns automatic adaptation.
- Keep Kinesis Provisioned mode but never change shards, because it is identical to On-demand scaling for variable traffic.
- Use one constant tenant partition key with On-demand mode because automatic capacity removes all hot-key per-shard limits.
- Send events only to SNS FIFO and assume it provides Kinesis Data Streams replay, retention, and shard semantics unchanged.
Explanation
Correct Answer
A:On-demand Standard adjusts stream capacity without manual shard planning and is designed for variable traffic; it still has growth and partition-key constraints.
Incorrect Answers
B:Provisioned mode requires planning/managing shard capacity and does not become automatic solely by keeping a fixed count.
C:A single hot partition key can still throttle a shard; On-demand scaling does not remove all partition-key distribution limits.
D:SNS FIFO is a messaging/pub-sub service with different durability and replay interface characteristics from Kinesis Data Streams.
Question 798
A sensor fleet sends hundreds of very small records per second into a provisioned Kinesis Data Streams shard. The service reaches its per-shard record-count limit well before the byte-per-second limit, although the consumer supports KPL deaggregation. The team wants higher effective user-record throughput without immediately adding shards. Which producer feature directly addresses the binding limit?
- Enable only PutRecords collection to put more separate Kinesis records into one HTTP request, assuming it lowers the number of stream records counted against the shard limit.
- Enable Kinesis Producer Library record aggregation to pack multiple user records into each Kinesis record, while monitoring the byte throughput and deaggregation requirements.
- Increase Kinesis stream retention from one day to seven days; the longer replay window expands per-shard write-records-per-second capacity.
- Register enhanced fan-out consumers, which increases producer-side shard record-per-second limits while leaving producer request patterns unchanged.
Explanation
Correct Answer
B:KPL aggregation packs several small user records into one Kinesis Data Streams record, reducing the counted service-record rate. Consumers must handle deaggregation and total byte/partition throughput remains bounded.
Incorrect Answers
A:PutRecords collection reduces producer HTTP request overhead but still sends separate Kinesis service records, so it does not by itself solve a binding per-shard record-count quota.
C:Retention changes how long records remain available for replay. It does not increase write-request capacity or the number of records a shard accepts per second.
D:Enhanced fan-out grants registered consumers dedicated read throughput. It does not change the producer-side shard write-record limit addressed by aggregation.
Question 799
A Firehose delivery stream writes telemetry into S3 in large batches, and product managers complain that dashboards can lag by several minutes. They want faster objects but understand small deliveries can raise request overhead. Which knobs control the tradeoff?
- Assume a Firehose S3 delivery stream creates one finished S3 object synchronously for every incoming record.
- Increase the data stream's retention to one year as the only fix for object-creation latency.
- Tune supported Firehose buffering size and interval settings for the destination and measure freshness versus delivery efficiency.
- Move the S3 destination bucket to Glacier Deep Archive before delivery to make objects arrive sooner.
Explanation
Correct Answer
C:Managed delivery buffers data until configured size/time thresholds, trading object size and ingest efficiency for latency.
Incorrect Answers
A:Buffering means object delivery need not be record-immediate.
B:Retention protects replay history, not S3 delivery buffer flushing.
D:Archival class changes do not reduce Firehose's batching interval.
Question 800
A Firehose pipeline receives JSON application logs, but the downstream Athena queries repeatedly scan raw uncompressed lines. The analytics team wants columnar Parquet files in S3. Which transformation should be evaluated?
- Enable only an ALB HTTP listener and expect it to convert JSON records into Parquet.
- Change object names from `.json` to `.parquet` without encoding the actual content.
- Increase Kinesis shard retention while leaving the Firehose delivery payload raw JSON.
- Configure supported Firehose data-format conversion or transformations and a valid Glue schema/catalog arrangement before S3 delivery.
Explanation
Correct Answer
D:Firehose can convert supported record formats for efficient analytic storage when schemas and destination compatibility are configured.
Incorrect Answers
A:A load balancer does not transform streaming payload formats.
B:A file extension does not make bytes conform to Parquet columnar format.
C:Retention does not transform the stored object serialization.
