Exam ANS-C00 topic 1 question 52 discussion

I think that since the questions notes that TCP and UDP are proprietary, any AWS objects that intercept the packet are not going to work - Cloudfront, ELB, etc. So you'll have to use DNS..answer is A.

Its A, ELB also does not support UDP , only NLB does

Weird set of options. Option B uses "ELB" term, which includes ALB, NLB, and CLB even. So, it does not rule out setting NLB to serve the proprietary TCP-UDP protocols. B is possibly right. A is sufficiently right. I'd go with A.

Correct is B. multiple NLBs (with TCP and UDP listener) can be used in different regions as part of a R53 latency-based routing scheme. Answer A was correct before NLB support for UDP.

Clearly it is B. A- Register IP address of every service hosts is not scalable. Context says " worldwide growth" B- Now NLB support UDP traffic. Then is CORRECT! C- CloudFront does not support UDP traffic. Is for HTTP. D- API GW is to disturb.

Hey guys 13 more new Qs, February 2022 added in Question 1

The question says "End users must have the lowest possible latency" removing the ELB from the equation improves latencies

In a real world situation, this is a good use case for global accelerator, however just to get through this question, I would go for A which is close. But the thing is that Apps sit in US EAST 1 and even if the route 53 latency based policy is used, it wont be that effective, unless APPs deployed in multiple regions. C is another best option, if this was HTTP/S traffic.

For all who say "not A, because route 53 does not support UDP" : we do not care We have a service, running on a particular IP addr, through both UDP and TCP We have a zone, with a A record, pointing to that IP In the zone, we have a check, unrelated to the actual service, that tells us if the host is healthy or not We could use the actually service as a check, as it is done to HTTP (make an http request to the service etc) We could also make a completly differente request, to another port and using another protocol, for instance Route 53 checks serves a single purpose: tell if the host is healthy How that check is implemented has nothing to do we the actual service

Route 53 health check doesn't support UDP... A is incorrect in that sense. I am more leaning towards B

Option A is correct. Altough NLB now support UDP it wasn't when this exam was created: NLB supports UDP since Jun 24, 2019. This exam is made in jan 2019 (ANS-C00).

B is correct

Route53 can load balance traffic globally (different region) but ELB cannot. Since, application/system is going to expand globally, best choice is A. Also, adding ELB as extra layer for inter-region will add latency, which is not lowest possible.

B is the right answer

Difficult to decide between A and B. Can someone elaborate what is meant with 'register the ELB name of the main service host as an ALIAS record' in C? One would use DNS name of ELB as ALIAS in R 53 or host directly, but what does ELB name of the main service mean?

B makes a lot of sense because it addresses the need for global expansion (or scalability) as well as low latency. The only sticking point is the proprietary TCP and UDP protocols. Not sure what does it imply here. A communication protocol is a system of rules that allow two or more entities of a communications system to transmit information. If it is a TCP and UDP implementation, it needs to follow the TCP and UDP rules. Anyway, this is where the confusion is.

A , based on proprietary TCP and UDP protocols for ELB.