Exam 350-901 All Questions

View all questions & answers for the 350-901 exam

Exam 350-901 topic 1 question 255 discussion

Actual exam question from Cisco's 350-901

Question #: 255
Topic #: 1

Refer to the exhibit. A Docker swarm service is currently running in a local data center. The service is hosting an HTML website. If the container fails, then the service becomes unavailable. The design must meet these requirements:
• The service must be highly available and resilient against a data center outage.
• The service must be accessible from a single URL.
• The HTTP session must remain on the server from which the original request was sent.
• Failure of the server must force the client to reconnect.

Which two design approaches must be used to meet the requirements? (Choose two.)

A. Create another node in the swarm cluster to scale the service across the nodes over two replicas.
B. Scale the Docker swarm service to 2 and set endpoint-mode to DNSRR instead of the default value of VIP.
C. Configure an external load balancer to route requests to the swarm service by using session persistence.
D. Configure a routing mesh to route requests to the swarm service by using NAT on the network side.
E. Create another swarm cluster within a data center and deploy a secondary instance of the service.

Show Suggested Answer

Suggested Answer: AC 🗳️

by fb48 at Jan. 14, 2023, 8:07 a.m.

Comments

Submit Cancel

fb48

Highly Voted 2 years, 5 months ago

Not sure, motivation: 4 requirements: A:only high avail part of 1 but not DC outage + 4 B:req 2 + 4 + high avail part of 1 not DC outage C:req high avail part of 1 + 2 +3 D:? E:req 1 + 4 My answer would be C+E

upvoted 9 times

lznlxl

2 years, 5 months ago

Agree with your answer. CE

upvoted 2 times

it0

2 years, 5 months ago

E would not protect against a data center outage

upvoted 4 times

...

Daimon336

Most Recent 7 months, 2 weeks ago

Selected Answer: AC

A - By creating another node in the Docker Swarm cluster and scaling the service to run on multiple nodes, the application becomes highly available. This setup means if one container or node fails, another replica will be available to handle requests, increasing resilience.

upvoted 2 times

...

Fedesarucho

1 year, 3 months ago

Selected Answer: CE

C and E as fb48 explains for disaster recovery requirement

upvoted 1 times

...

Alexissiox

2 years, 1 month ago

The service must be accessible from a single URL. Cannot be B since the endpoint-mode is set to DNSRR... A&C

upvoted 3 times

...

cj_kuo

2 years, 2 months ago

Selected Answer: AC

I agree the answer as A and C, since the answer from a friend who passed the exam.

upvoted 2 times

...

[Removed]

2 years, 5 months ago

I asked ChatGPT and it gave this answer: A & C How confident are you in your answer? I am highly confident in my answer, the options A and C are the best fit to meet the requirements, Creating another node in the swarm cluster to scale the service across the nodes over two replicas ensures high availability and resiliency against a data center outage. Configuring an external load balancer to route requests to the swarm service by using session persistence ensures that the service is accessible from a single URL and that the HTTP session remains on the server from which the original request was sent. It also makes sure that failure of the server forces the client to reconnect.

upvoted 2 times

Fedesarucho

2 years, 1 month ago

B. Scale the Docker swarm service to 2 and set endpoint-mode to DNSRR instead of the default value of VIP. C. Configure an external load balancer to route requests to the swarm service by using session persistence. These two design approaches can help meet the requirements mentioned. Scaling the Docker swarm service to two replicas allows for high availability and resilience against a data center outage. Setting endpoint-mode to DNSRR ensures that requests are distributed across replicas in a round-robin fashion. Configuring an external load balancer to route requests to the swarm service with session persistence will ensure that HTTP sessions remain on the server from which the original request was sent, and if a server fails, the client will be forced to reconnect. now it changed its opinion :/

upvoted 1 times

...