01. A ticketing platform's RDS for MySQL database must keep serving orders automatically if the primary instance's AZ fails. Separately, heavy sales reports run during the day are slowing order writes.
Which two measures address these requirements? (Choose two.)
a) Send the reports to the standby instance
b) Use a primary/standby RDS instance with the standby in a different AZ
c) Increase the automated backup frequency so the database can be restored after an AZ failure
d) Add read replicas and direct the sales reports to them
e) Scale up the primary instance's vCPUs and memory to absorb both the reports and an AZ outage
02. A regional newspaper's website already uses Anti-DDoS, which absorbed a traffic flood last month. This week, access logs show URL parameters carrying SQL fragments, and one web ECS is running an unknown cryptocurrency-mining process that connects out to unfamiliar addresses.
Which additions address the two new findings?
a) WAF to filter the injection attempts, and HSS to detect and handle the mining process on the server
b) A stricter Anti-DDoS policy for the injection attempts, and HSS for the mining process
c) WAF for both findings, since the miner was most likely installed through HTTP requests
d) HSS to filter the injection attempts in web traffic, and WAF to detect and handle the mining process on the server
03. A sports league's web tier runs in an AS group in one AZ. The league now wants the group to spread instances across two AZs and never drop below three instances. The current flavor and image must stay unchanged.
Which two changes should be made on the AS group itself? (Choose two.)
a) Create a new AS configuration whose minimum is 3 instances
b) Add the second AZ to the AS configuration
c) Set the AS group's minimum number of instances to 3
d) Attach an alarm policy that keeps at least three instances running
e) Add the second AZ in the AS group settings
04. A ticketing platform's AS group adds instances through an alarm policy on CPU usage. Load arrives at unpredictable times, and a new instance needs about six minutes before it starts taking traffic. During each surge the group keeps adding instances well past what the load needs, then removes them again.
Which change addresses the overshoot while still reacting to unpredictable load?
a) Lower the CPU threshold of the alarm policy so that scaling starts earlier in each surge
b) Set a cooldown period longer than the warm-up time, so new instances absorb load before the group acts again
c) Replace the alarm policy with a scheduled policy that adds instances at the busiest hours
d) Reduce the AS group maximum to the number of instances the largest surge has needed
05. A genomics lab keeps its ECSs in a private subnet with no EIPs bound to them. Every ECS must download OS updates from the internet, and one ECS hosts a results API that external partners must reach on a single port. The ECSs must stay without public addresses of their own.
Which two NAT Gateway rules should be configured? (Choose two.)
a) An SNAT rule for the private subnet, using an EIP for outbound traffic
b) A DNAT rule for the private subnet, so every ECS can reach the update servers
c) An SNAT rule on the API server, so partners can open connections to it
d) A DNAT rule mapping an EIP and port to the API server's private address and port
e) An EIP bound to each ECS, keeping the NAT Gateway only as a fallback path
06. A university's research subnet has a network ACL with two inbound rules. The rule evaluated first allows TCP 443 from 0.0.0.0/0. The rule evaluated second denies all traffic from 198.51.100.0/24. Hosts in 198.51.100.0/24 still reach the HTTPS service.
What should the administrator change?
a) Narrow the deny rule to a single host address so it becomes the more specific match
b) Change the deny rule to match TCP 443 only, so it covers exactly the same traffic
c) Reorder the rules so the deny for the range is evaluated before the allow
d) Associate the ACL with the subnet again so the second rule takes effect
07. An architecture firm keeps project drawings in an OBS bucket. Last month a faulty script overwrote and deleted hundreds of drawings. Management also wants the drawings to survive if the bucket's whole region becomes unavailable.
Which measures address these two risks? (Choose two.)
a) Enable bucket encryption with a KMS key so that a deleted object can be recovered through its key
b) Enable versioning on the bucket so that overwritten or deleted drawings keep their earlier versions
c) Change the drawings to the Archive class so that a stray script cannot overwrite them
d) Turn on access logging for the bucket so deleted drawings can be rebuilt from the logs
e) Configure cross-region replication to a bucket in another region
08. A streaming service serves /video/ requests from one group of ECSs and /api/ requests from another, both under one HTTPS domain on one load balancer. The operations team also wants TLS decrypted once at the load balancer rather than on every ECS.
Which ELB configuration meets both requirements?
a) An HTTPS listener whose forwarding policies match the URL path of a request
b) A TCP listener with source IP hash, so that each client keeps reaching its group
c) Two load balancers with one EIP each, chosen by clients according to the path
d) A TCP listener on port 443 with a backend server group per path
09. A university is attaching two EVS data disks to a research ECS. One holds a database with heavy random reads and writes that must respond with low latency. The other stores weekly log archives that are written once and rarely read, and its budget is tight. The available disk types differ in performance and price.
Which disk types fit these two needs?
a) A faster, higher-priced type for both disks, to keep their performance uniform
b) A cheaper type for the database, and a faster, higher-priced type for the archives
c) A cheaper type for both disks, to keep the total cost as low as possible
d) A faster, higher-priced type for the database, and a cheaper type for the archives
10. A bank's back office runs a nightly reconciliation script that calls HUAWEI CLOUD APIs as an IAM user. The script's access key has just been found in a public code repository. Separately, auditors require a second verification factor whenever an administrator signs in to the console.
Which pair of actions addresses both findings?
a) Enable MFA on the script's IAM user so that each API call needs a code, and keep the existing AK/SK
b) Change the script user's console password, and have the administrators change their passwords every month
c) Replace the script's leaked AK/SK, and enable MFA login protection for the administrators
d) Delete the administrators' access keys, and let the script sign its API calls with the account's own credentials