Question 1

An SageMaker inference endpoint is ...

Accepted Answer

a network-connected deployment of a machine learning model that can be used for live classification/prediction of new data records.

Answer

a deployment of a machine learning model on your local machine that can be used for live classification/prediction of new data records.

Answer

a browser-based IDE for machine learning tasks.

Answer

an extension to the Python boto3 library that can be used to run a model for live classification/prediction of new data records.

Question 2

A model deployed via a SageMaker inference endpoint is available continuously and awaiting requests for classification/prediction.

Accepted Answer

true

Answer

false

Question 3

Select an advantage of running a machine learning IDE (e.g., SageMaker) in the cloud, instead of your local machine.

Accepted Answer

It allows to directly work with data that may already be stored in the cloud ("data locality").

Answer

It's cheaper than running on a local machine.

Answer

Machine learning experts employed by the cloud service provider can have a look at your machine learning notebooks and bring in suggestions for improvement.

Question 4

What is a potential issue when writing infrastructure-as-code templates in the form of regular shell scripts?

Accepted Answer

Sending API calls directly, like done via a shell script, often only handles the "create case" of infrastructure. If infrastructure is updated or deleted, additional script logic needs to be written.

Answer

There is only limited support from cloud service providers to send API commands directly from the shell.

Answer

There are maximum execution times associated with shell scripts that can be hit when deploying large infrastructure.

Question 5

Infrastructure-as-code tools often follow a "desired state model" approach to define and manage infrastructure.

Accepted Answer

true

Answer

false

Question 6

The "Resource Manager" service, which allows to follow an infrastructure-as-code modeling approach, belongs to the following cloud service provider.

Accepted Answer

Azure

Answer

Alibaba

Answer

GCP

Answer

AWS

Question 7

It is possible to deploy the very same Terraform code definition to multiple cloud service providers (e.g., AWS and Azure), which reduces development effort in multi-cloud solutions.

Accepted Answer

false

Answer

true

Question 8

Terraform can be used to deploy to multiple cloud service providers at the same time (i.e., during the same "terraform apply" command). Terraform will figure out which "resource" blocks in the *.tf files belong to which cloud service provider and will deploy the resources accordingly.

Accepted Answer

true

Answer

false

Related quizzes

Related quizzes