350-901 question 30: Which two actions are required when developing an observable micros…

Q30Application Deployment and SecurityMultiple answers

Which two actions are required when developing an observable microservice application? (Choose two.)

Select 2 answers.

← → navigate · a answer

Community votes

50% (4)

0% (0)

Discussion · 11

I’m choosing CD

I’ll go with C and D based on: https://opensource.com/article/18/9/distributed-tracing-microservices-world

C, D 4

Selected Answer: CD Correct answers: C & D Logging: Logging keeps track of events and their timestamps. Logs and log types matter too. Tracing: Tracing is the ability to follow multiple events or a chain of distributed events through a system. https://opensource.com/article/18/9/distributed-tracing-microservices-world

C, D 3

Selected Answer: CD C and D are correct: Observability has three pillars (Logging, Metrics, and Tracing) or telemetry types that you, as application developer or designer, need to consider: Logging: Logging tracks events and their timestamps. Logs and log types are also important. Metrics or time-series metrics: These metrics can simply be defined as system performance parameters (or application health measures) and are usually measured within a unit in time. As mentioned in “Latency and Rate Limiting in Application Design and Performance,” response time, sessions per second, transactions per second—all are examples of metrics. Tracing: Tracing is the ability to track multiple events or a series of distributed events through a system.

"B" is not part of "observable microservice" gonna go with CD

I agree with C and D!

Why do I think AD is better? What’s wrong with A?

C, D 2

Selected Answer: CD Should be C&D

I lean toward C&D because that’s what you do during development. Monitoring containers matters too, but that is Ops and is often provided by the platform.

The answer is in the url provided by B3nd2cida "knowing the state of a single instance of a single service isn't enough; we need to be able to aggregate the metrics for all instances of a given service, perhaps grouped by version"

"knowing the state of a single instance of a single service isn't enough". Yes, that is true. But not being enough does not mean it should not be part of a solution. It just means the solution is multi-factor. Several things need to be in place and none of them alone is enough. To me this is just another classic example of Cisco's long tradition of making messed-up questions that can't be answered 100% correctly. In this question, the right answer to me would be A, C and D.