Cuestionario vinculado a una guía · Medio Nivel

Prueba de Terraform para infraestructura de aprendizaje automático

Revise la configuración, los planes y el estado de Terraform, prestando atención a la capacidad de la GPU, la seguridad del estado remoto y los cambios de producción controlados.

Rutas de guía relacionadasTerraform para infraestructura de aprendizaje automático
Pregunta 1 de 8

What does Terraform state primarily track?

Vista previa de cada pregunta8 preguntas

Las respuestas y explicaciones aparecen a medida que realiza el cuestionario.

  1. What does Terraform state primarily track?
    • The relationship between declared resources and remote infrastructure
    • The model's validation accuracy
    • The latest GPU prediction output
    • The Python package graph
  2. Which review step shows proposed infrastructure actions before they are applied?
    • Guarantee cloud capacity at apply time
    • Inspect proposed infrastructure changes before applying them
    • Validate model fairness automatically
    • Replace remote state
  3. Why protect remote Terraform state?
    • It guarantees GPU availability
    • It contains only public source code
    • It can contain sensitive values and is needed to manage resources reliably
    • It replaces provider credentials
  4. Why might a valid GPU plan still fail during apply?
    • State automatically creates cloud stock
    • Terraform plans are always stale by definition
    • The GPU model determines Python syntax
    • Regional capacity or account quota may be unavailable
  5. What does setting a variable as sensitive guarantee?
    • That the secret is encrypted in every backend
    • Display redaction, not automatic removal from state
    • That no provider can read it
    • That it is omitted from all logs and files
  6. Why separate model validation from infrastructure provisioning?
    • Model evaluation changes cloud regions
    • Terraform cannot reference container images
    • Terraform can create resources without determining whether a model is fit to deploy
    • Infrastructure plans contain no resources
  7. What should a team do with a plan that includes destructive production actions?
    • Remove backups first
    • Apply automatically because Terraform calculated it
    • Ignore state and repeat the apply
    • Inspect affected resources, dependencies and recovery needs before applying
  8. Why use narrow apply permissions in production?
    • It limits who can make infrastructure changes
    • It makes every plan smaller
    • It prevents all provider outages
    • It removes the need for source review