Question 440
At Granite SaaS, the site reliability team maintains multi-tenant Linux application instances under requirements for per-service isolation and reliable automated scaling. The operations lead requests proof of the fix for: A newly provisioned cloud VM needs a one-time user and package bootstrap but operators manually SSH to every instance after launch. Which observed result BEST demonstrates that the correction works?
Explanation
Correct Answer
A fresh test instance has the intended account and packages immediately after cloud-init completes without ad hoc operator login. This measurable outcome verifies correction of: A newly provisioned cloud VM needs a one-time user and package bootstrap but operators manually SSH to every instance after launch. The associated technical action is: Use a reviewed cloud-init configuration for first-boot setup and preserve the metadata and secret-handling boundaries. The Granite SaaS review checks actual behavior rather than an untested change claim.
Incorrect Answers
The change appears as an attributable reviewed commit and the protected branch rejects an unapproved direct push. This evidence pertains to an unrelated remediation: Create a feature branch with a focused commit and submit a reviewed merge process before updating the protected main branch. The alternative issue is: A team edits configuration directly on the main Git branch without preserving a reviewable history of the proposed change. The current Granite SaaS problem is: A newly provisioned cloud VM needs a one-time user and package bootstrap but operators manually SSH to every instance after launch. The decisive expected evidence is instead: A fresh test instance has the intended account and packages immediately after cloud-init completes without ad hoc operator login.
Boundary inputs such as 9, 10 and 100 produce the correct threshold decisions without lexicographic errors. This evidence pertains to an unrelated remediation: Use the supported numeric comparison operators for integer test expressions and validate the input before comparison. The alternative issue is: A data-cleanup script performs arithmetic comparison of a numeric threshold using string ordering and misclassifies double-digit values. The current Granite SaaS problem is: A newly provisioned cloud VM needs a one-time user and package bootstrap but operators manually SSH to every instance after launch. The decisive expected evidence is instead: A fresh test instance has the intended account and packages immediately after cloud-init completes without ad hoc operator login.
The release tag resolves to the approved tested commit and an unreviewed commit cannot receive the protected release tag. This evidence pertains to an unrelated remediation: Create a protected tag only after the tested commit hash passes its release approval and provenance checks. The alternative issue is: A release tag is created before required tests finish, causing automation to deploy an unverified commit as a production version. The current Granite SaaS problem is: A newly provisioned cloud VM needs a one-time user and package bootstrap but operators manually SSH to every instance after launch. The decisive expected evidence is instead: A fresh test instance has the intended account and packages immediately after cloud-init completes without ad hoc operator login.