SAP-C03 · Domain 1: Architecture Design and Implementation Built for the Cloud · 25% of the exam
Task 1.4: Design network architectures.
Many-VPC networks: peering, PrivateLink and transit routing, IP address planning with IPAM, inspection and filtering with Network Firewall, WAF, Gateway Load Balancer and flow logs, IPv6 and dual stack, and global routing with Route 53 and Global Accelerator.
Study it
Many-VPC networks: peering, PrivateLink, Transit Gateway, Cloud WAN and IP address planning
Try each one before opening the answer. Every option is explained, with the AWS documentation page that proves it.
Question 1 · choose 1
A company has 45 VPCs in 12 accounts of its organization, split between us-east-1 and eu-west-1. Every VPC must be able to reach every other VPC, and the on-premises data center, connected through one AWS Direct Connect connection in us-east-1, must reach the VPCs in both Regions. The network team wants the fewest connections to manage and wants to add VPCs later without touching the existing ones. Which design meets these requirements?
APeer every VPC with every other VPC, and create a private virtual interface from the Direct Connect connection to a virtual private gateway on each VPC
BUse a transit gateway per Region with the local VPCs attached, peer the two, and associate both with one Direct Connect gateway over a transit VIF
CCreate one transit gateway in us-east-1, attach the VPCs of both Regions to it, and connect the Direct Connect connection to it through a transit virtual interface
DAttach a virtual private gateway to each VPC, associate all of them with one Direct Connect gateway, and let the Direct Connect gateway route traffic between the VPCs
Show the answer and why
APeer every VPC with every other VPC, and create a private virtual interface from the Direct Connect connection to a virtual private gateway on each VPC
Incorrect
VPC peering is one-to-one and not transitive, so 45 VPCs need a full mesh of peering connections, and every new VPC adds a peering connection to each existing VPC. That is the opposite of fewest connections.
BUse a transit gateway per Region with the local VPCs attached, peer the two, and associate both with one Direct Connect gateway over a transit VIF
Correct
A transit gateway is a Regional router for the VPCs attached to it, and transit gateways in different Regions can be peered. A Direct Connect gateway associated with transit gateways through a transit virtual interface gives on premises a path to their attached VPCs.
CCreate one transit gateway in us-east-1, attach the VPCs of both Regions to it, and connect the Direct Connect connection to it through a transit virtual interface
Incorrect
A transit gateway acts as a Regional virtual router. VPCs in another Region are connected through a transit gateway in their own Region that is peered with the first one.
DAttach a virtual private gateway to each VPC, associate all of them with one Direct Connect gateway, and let the Direct Connect gateway route traffic between the VPCs
Incorrect
A Direct Connect gateway does not let the gateways associated with it send traffic to each other, so it cannot provide VPC-to-VPC routing between virtual private gateways.
Many VPCs in several Regions plus a hybrid link is the transit gateway pattern: one transit gateway per Region, peering between them, and a Direct Connect gateway associated with the transit gateways. New VPCs only need an attachment to their Regional transit gateway.
After a network change, application servers in one member account can no longer open TCP connections on port 5432 to an Amazon RDS for PostgreSQL database in another member account. Both VPCs are attached to the same transit gateway. The network team must name the configuration item that blocks the path and must confirm whether the packets reach the database's network interface at all. Which actions should the team take? (Choose TWO.)
ARun a Reachability Analyzer path analysis from the application instance to the database interface, with organization trusted access
BTurn on AWS CloudTrail data events for Amazon RDS and search them for rejected connections from the application subnets
CTurn on RDS Enhanced Monitoring and look for dropped connections in the operating system metrics of the database host instance
DPublish VPC Flow Logs for the database subnet and look for ACCEPT or REJECT records from the application CIDR range on port 5432
ETurn on AWS X-Ray active tracing for the database so that each hop between the two VPCs appears in the trace map
Show the answer and why
ARun a Reachability Analyzer path analysis from the application instance to the database interface, with organization trusted access
Correct
When the destination is not reachable, Reachability Analyzer names the blocking component, such as a security group, network ACL or route table. With trusted access in Organizations it analyzes paths across accounts.
BTurn on AWS CloudTrail data events for Amazon RDS and search them for rejected connections from the application subnets
Incorrect
CloudTrail records API activity. A TCP connection to a database port is network traffic, not an API call, so it does not appear in CloudTrail.
CTurn on RDS Enhanced Monitoring and look for dropped connections in the operating system metrics of the database host instance
Incorrect
Enhanced Monitoring reports operating system metrics such as CPU and memory for the DB instance. It does not show which network component dropped a packet before it reached the instance.
DPublish VPC Flow Logs for the database subnet and look for ACCEPT or REJECT records from the application CIDR range on port 5432
Correct
Flow log records show the source, destination, port and an ACCEPT or REJECT action for traffic at the network interface, which confirms whether packets arrive and whether they are rejected.
ETurn on AWS X-Ray active tracing for the database so that each hop between the two VPCs appears in the trace map
Incorrect
X-Ray traces requests through instrumented application code. It does not trace packets through route tables, network ACLs or transit gateways.
Reachability Analyzer explains a path from configuration without sending traffic and names the blocker; flow logs show what actually arrived at the interface. CloudTrail, Enhanced Monitoring and X-Ray look at other layers.
A software-as-a-service provider runs a TCP service behind a Network Load Balancer in its own VPC. About 300 customers in their own AWS accounts must reach it privately, without traffic crossing the internet. Many customer VPC CIDR ranges overlap with each other and with the provider's VPC. Customers must be able to start connections to the service, but the provider must not get a network path into customer VPCs. Which solution meets these requirements?
AExpose the Network Load Balancer as a VPC endpoint service and have each allowed customer create an interface VPC endpoint
BShare a transit gateway with the customer accounts through AWS RAM and attach every customer VPC to it with routes to the provider VPC
CCreate a VPC peering connection between the provider VPC and each customer VPC, and add routes only for the load balancer subnets
DAsk each customer to build an AWS Site-to-Site VPN connection from its VPC to a virtual private gateway on the provider VPC
Show the answer and why
AExpose the Network Load Balancer as a VPC endpoint service and have each allowed customer create an interface VPC endpoint
Correct
With AWS PrivateLink the consumer creates an interface endpoint and starts the connection, the provider allows chosen principals, and the endpoint's network interfaces sit in the consumer VPC, so overlapping address ranges do not conflict.
BShare a transit gateway with the customer accounts through AWS RAM and attach every customer VPC to it with routes to the provider VPC
Incorrect
A transit gateway routes by destination address, so overlapping CIDR ranges cannot be told apart, and routed attachments give network paths in both directions rather than one-way access to a single service.
CCreate a VPC peering connection between the provider VPC and each customer VPC, and add routes only for the load balancer subnets
Incorrect
AWS does not allow a VPC peering connection between VPCs with matching or overlapping CIDR blocks, so many of these customers could not peer at all.
DAsk each customer to build an AWS Site-to-Site VPN connection from its VPC to a virtual private gateway on the provider VPC
Incorrect
VPN connections would still route between overlapping address ranges, and each one gives a routed network path between the two networks, which the provider must avoid.
PrivateLink exposes one service, not a network: the consumer connects to an endpoint inside its own VPC, the provider accepts it, and address overlap does not matter.
A security team found malware that exfiltrated data by encoding it in DNS queries to an attacker's domains from EC2 instances in several VPCs. It wants to block queries to known malicious domains and allow only approved domains for some workloads, using the VPCs' default DNS resolution. Which solution meets these requirements?
AUse security group outbound rules that block UDP port 53 to the internet for all instances
BAttach an AWS WAF web ACL to each VPC to filter DNS requests
CAnalyze VPC Flow Logs every night for unusual DNS traffic patterns
DAssociate Route 53 Resolver DNS Firewall rule groups with block and allow lists
Show the answer and why
AUse security group outbound rules that block UDP port 53 to the internet for all instances
Incorrect
Instances resolve names through the VPC resolver, not directly over the internet, so this does not filter which domains are resolved.
BAttach an AWS WAF web ACL to each VPC to filter DNS requests
Incorrect
AWS WAF protects web applications behind supported resources; it does not attach to VPCs or filter DNS queries.
CAnalyze VPC Flow Logs every night for unusual DNS traffic patterns
Incorrect
Flow logs record traffic metadata after the fact and do not block queries.
DAssociate Route 53 Resolver DNS Firewall rule groups with block and allow lists
Correct
DNS Firewall filters outbound DNS queries that go through the VPC resolver, and a primary use is to help prevent DNS exfiltration.
Domain-based control of outbound DNS from VPCs is Route 53 Resolver DNS Firewall.
Sixty VPCs attach to one transit gateway. Production VPCs and development VPCs must not reach each other, but both must reach a shared services VPC and the on-premises network. Which transit gateway design meets this?
ANetwork ACLs in the shared services VPC that block development traffic
BA separate transit gateway for each VPC
CSeparate route tables per environment, propagating only shared services and on-premises routes
DOne default route table for all attachments, with security groups in every VPC to block traffic
Show the answer and why
ANetwork ACLs in the shared services VPC that block development traffic
Incorrect
Network ACLs in one VPC do not stop production and development VPCs from routing to each other.
BA separate transit gateway for each VPC
Incorrect
Sixty transit gateways multiply cost and routing work.
CSeparate route tables per environment, propagating only shared services and on-premises routes
Correct
Attachments are associated with route tables, and routes are propagated from chosen attachments, so each environment learns only the routes it should reach.
DOne default route table for all attachments, with security groups in every VPC to block traffic
Incorrect
A single route table gives every attachment routes to every other one, leaving isolation to many security groups.
Segmenting traffic on a transit gateway is done with separate route tables, associations and propagations.
A retailer is launching a global storefront on Amazon CloudFront with an Application Load Balancer origin. It expects large HTTP request floods during sales events. The company wants help from AWS experts during an attack and protection from the extra AWS charges that an attack could cause by scaling the application. Which solution meets these requirements?
ARely on AWS Shield Standard and add Amazon GuardDuty to detect attacks against the load balancer and the distribution
BSubscribe to AWS Shield Advanced for the distribution and load balancer, and add AWS WAF rate-based rules
CPut AWS Network Firewall in the origin VPC and drop traffic from source IP addresses that send too many requests per second
DMove the origin to a larger instance type and raise the Auto Scaling group's maximum size for sales events
Show the answer and why
ARely on AWS Shield Standard and add Amazon GuardDuty to detect attacks against the load balancer and the distribution
Incorrect
Shield Standard is automatic and free but gives no access to the Shield Response Team or cost protection. GuardDuty detects threats; it does not block request floods.
BSubscribe to AWS Shield Advanced for the distribution and load balancer, and add AWS WAF rate-based rules
Correct
Shield Advanced adds access to the Shield Response Team and cost protection for scaling charges caused by a DDoS attack, and AWS WAF rate-based rules limit floods of web requests.
CPut AWS Network Firewall in the origin VPC and drop traffic from source IP addresses that send too many requests per second
Incorrect
Network Firewall inspects traffic in a VPC, behind the edge. It offers neither the Shield Response Team nor cost protection.
DMove the origin to a larger instance type and raise the Auto Scaling group's maximum size for sales events
Incorrect
More capacity absorbs some load but increases exactly the attack-driven charges the company wants protection from.
Expert help during an attack and protection from attack-driven charges are Shield Advanced features; WAF rate-based rules handle the request floods.
A web application will run in eu-central-1, us-east-1 and ap-southeast-1. For legal reasons, users in Germany must always be served from eu-central-1. All other users must be served from the Region with the lowest latency for them. Which Route 53 configuration meets these requirements?
ALatency records for the three Regions, each with a health check so that unhealthy Regions are skipped automatically
BGeoproximity records with a positive bias on eu-central-1 large enough to cover Germany
CA geolocation record for Germany to eu-central-1, plus a default geolocation record that aliases latency records
DWeighted records that send a share of traffic to each Region in proportion to its share of users
Show the answer and why
ALatency records for the three Regions, each with a health check so that unhealthy Regions are skipped automatically
Incorrect
Latency routing picks the Region with the lowest latency, which for some German users might not be eu-central-1, so the legal requirement is not guaranteed.
BGeoproximity records with a positive bias on eu-central-1 large enough to cover Germany
Incorrect
Geoproximity routing shifts traffic by distance and bias. It does not guarantee that every user in one country goes to one Region.
CA geolocation record for Germany to eu-central-1, plus a default geolocation record that aliases latency records
Correct
Geolocation routing answers by the user's location, and the default record catches every other location. Records with different routing policies can be combined by aliasing one to the other.
DWeighted records that send a share of traffic to each Region in proportion to its share of users
Incorrect
Weighted routing distributes traffic by weight regardless of the user's location or latency.
Hard location rules use geolocation; "everyone else by latency" sits behind the default geolocation record as a latency record set.
A multiplayer game runs its UDP game servers on EC2 instances in two Regions. Players around the world complain about jitter on long internet paths, and some of their corporate firewalls only allow traffic to a fixed list of IP addresses. Which change improves the experience and meets the firewall constraint?
APut an AWS Global Accelerator standard accelerator in front of the game servers in both Regions
BServe the game traffic through an Amazon CloudFront distribution with the game servers as origins
CUse Route 53 latency-based records that point players to the Region with the lowest latency
DMove the game servers to larger instances with enhanced networking in both Regions
Show the answer and why
APut an AWS Global Accelerator standard accelerator in front of the game servers in both Regions
Correct
Global Accelerator provides static anycast IP addresses and improves TCP and UDP applications by carrying their traffic over the AWS global network to endpoints in the nearest Region.
BServe the game traffic through an Amazon CloudFront distribution with the game servers as origins
Incorrect
CloudFront speeds up cacheable and dynamic web content. For non-HTTP use cases such as UDP gaming, AWS points to Global Accelerator instead.
CUse Route 53 latency-based records that point players to the Region with the lowest latency
Incorrect
Latency records pick a Region at DNS time, but traffic still crosses the public internet, and the addresses are not a small fixed list.
DMove the game servers to larger instances with enhanced networking in both Regions
Incorrect
Enhanced networking speeds up the instance network interface. It does not shorten the internet path that causes the jitter.
Non-HTTP traffic, a global audience and fixed IP addresses point to Global Accelerator.