fix(server): warn on dangling provider refs at startup

[ericcurtin]

Summary

A deleted provider no longer blocks gateway startup. The preflight warns and keeps validating.

Related Issue

Closes #4657

Changes

• Preflight looks up each provider and warns per sandbox on missing ones.
• Split load_provider_environment_record out of load_provider_environment_records to share the lookup.

Testing

• Checks appropriate to the affected code and behavior pass
• Unit tests added/updated (if applicable)
• E2E tests added/updated (if applicable)

Ran cargo clippy and the startup_preflight unit tests.

Checklist

• Follows Conventional Commits
• Commits are signed off (DCO)
• Architecture docs updated (if applicable)

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[ericcurtin]

@johntmyers @elezar PTAL, and /ok to test 1ed3a28e09724eae20ff8ed886ff1cf641858ef3 if it looks good. Thank you!

[ericcurtin]

Merged main to fix conflicts. @johntmyers @elezar PTAL, and /ok to test 30c1f3b952de6bc334f529ed0338470e4daff6da if it looks good. Thank you!

[mavromixaliskoullis-beep]

Is ok

[mrunalp]

/ok to test 30c1f3b

[github-actions]

Label test:e2e applied, but pull-request/4661 does not exist yet. A maintainer needs to comment /ok to test 30c1f3b952de6bc334f529ed0338470e4daff6da to mirror this PR. Once the mirror exists, re-apply the label or re-run Branch E2E Checks from the Actions tab.

[ericcurtin]

E2E failures look unrelated to this change: k3s-ha pod gating timeout, ssh_host_identity container exit, and workspace operator lbl-sb1 already exists. Could a maintainer re-run the failed jobs? Thank you!