fix(onboard): use 127.0.0.1 instead of localhost for local inference detection by dumko2001 · Pull Request #176 · NVIDIA/NemoClaw

dumko2001 · 2026-03-17T10:23:05Z

Rationale

Using localhost can sometimes resolve to IPv6 addresses or behave inconsistently across different environments and network configurations.

Changes

Switched to the explicit IPv4 loopback address 127.0.0.1 for local inference service detection to improve reliability.

Verification Results

Automated Tests: Passed all 52 core tests via npm test.
Manual Audit: Verified connection reliability with 127.0.0.1.
Security Review: Verified no sensitive data leaks and correct permission enforcement.

Leading Standards

This PR follows the project's 'First Principles' approach, prioritizing deterministic behavior and zero-trust security defaults.

Summary by CodeRabbit

Bug Fixes
- Fixed detection of local Ollama and vLLM instances to probe the correct loopback address.
- Updated onboarding UI option labels and console messages to display the accurate local endpoint addresses for local inference providers.

wscurran · 2026-03-19T22:50:30Z

I've looked at the changes and see that this PR updates the local inference detection to use 127.0.0.1 instead of localhost, which should improve reliability across different environments and network configurations.

cv · 2026-03-21T19:28:44Z

Hi @dumko2001! The localhost vs 127.0.0.1 distinction is one of those subtle things that can really bite you. We'd like to consider this — please rebase onto the latest main so we can give it a proper review. A lot has changed upstream.

…detection

coderabbitai · 2026-03-21T21:34:02Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 0a44c887-1c56-424c-bf35-c4b9c1bc7f03

📥 Commits

Reviewing files that changed from the base of the PR and between abc1036 and 9d2653c.

📒 Files selected for processing (5)

bin/lib/local-inference.js
bin/lib/nim.js
bin/lib/onboard.js
test/local-inference.test.js
test/onboard-selection.test.js

✅ Files skipped from review due to trivial changes (4)

bin/lib/nim.js
test/local-inference.test.js
test/onboard-selection.test.js
bin/lib/local-inference.js

🚧 Files skipped from review as they are similar to previous changes (1)

bin/lib/onboard.js

📝 Walkthrough

Walkthrough

Switched local inference health-check and probe endpoints from localhost to 127.0.0.1 across onboarding, local-inference, and NIM health-check logic; updated related UI labels, console messages, and tests to match the new loopback address.

Changes

Cohort / File(s)	Summary
Onboarding UI & probes `bin/lib/onboard.js`	Probe URLs changed from `http://localhost:<port>` to `http://127.0.0.1:<port>` for Ollama and vLLM; UI option labels and console messages updated to show `127.0.0.1:<port>` instead of `localhost:<port>`.
Local inference checks & commands `bin/lib/local-inference.js`	Health-check curl endpoints, warmup and probe shell commands updated to target `http://127.0.0.1:<port>` for Ollama and vLLM; error messages adjusted to reference `127.0.0.1`. Container reachability checks unchanged.
NIM health checks `bin/lib/nim.js`	Health-check calls for NIM models updated to use `http://127.0.0.1:${port}/v1/models` instead of `localhost`.
Tests updated to match loopback address `test/local-inference.test.js`, `test/onboard-selection.test.js`	Test expectations and mocked command outputs updated from `localhost` to `127.0.0.1` for Ollama (`11434`) and vLLM (`8000`) probes and warmup/probe commands.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Poem

🐰 I hopped the network, sniffed the log,
127.0.0.1 — a snug little hog,
Ollama hummed and vLLM replied,
Labels shifted, probes now tried,
A rabbit's cheer for loopback pride! 🥕

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 12.50% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and accurately summarizes the primary change: replacing localhost with 127.0.0.1 across local inference detection in onboard, NIM, and local-inference modules.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

coderabbitai

🧹 Nitpick comments (1)

bin/lib/onboard.js (1)

549-556: Unify endpoint wording across the onboarding flow.

Line 549 and Line 556 now show 127.0.0.1, but Line 652, Line 665, and Line 673 still print localhost. Keep these consistent to reduce troubleshooting ambiguity.

♻️ Suggested consistency patch

-      console.log("  ✓ Using Ollama on localhost:11434");
+      console.log("  ✓ Using Ollama on 127.0.0.1:11434");
...
-      console.log("  ✓ Using Ollama on localhost:11434");
+      console.log("  ✓ Using Ollama on 127.0.0.1:11434");
...
-      console.log("  ✓ Using existing vLLM on localhost:8000");
+      console.log("  ✓ Using existing vLLM on 127.0.0.1:8000");

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@bin/lib/onboard.js` around lines 549 - 556, The onboarding labels use mixed
hostnames ("127.0.0.1" in some labels and "localhost" in others); update the
other label strings to match the chosen canonical form (use "127.0.0.1" per
recent edits) so wording is consistent. Search for the label construction sites
where options are pushed (look for options.push and labels that mention
"localhost") and replace "localhost" with "127.0.0.1", preserving surrounding
text, experimental tags, and the suggested/running suffix logic (refer to the
ollamaRunning, vllmRunning and EXPERIMENTAL usage to ensure behavior and
conditional suffixes remain unchanged).

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@bin/lib/onboard.js`:
- Around line 549-556: The onboarding labels use mixed hostnames ("127.0.0.1" in
some labels and "localhost" in others); update the other label strings to match
the chosen canonical form (use "127.0.0.1" per recent edits) so wording is
consistent. Search for the label construction sites where options are pushed
(look for options.push and labels that mention "localhost") and replace
"localhost" with "127.0.0.1", preserving surrounding text, experimental tags,
and the suggested/running suffix logic (refer to the ollamaRunning, vllmRunning
and EXPERIMENTAL usage to ensure behavior and conditional suffixes remain
unchanged).

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: ce39aac5-cf17-4f4c-884d-5f93628e4341

📥 Commits

Reviewing files that changed from the base of the PR and between 255e606 and abc1036.

📒 Files selected for processing (1)

bin/lib/onboard.js

…x tests

…d policy identity clearing (NVIDIA#176)

brandonpelfrey

Thanks for the cleanup here. I found one blocking issue:

NIM path/tests are now inconsistent with the stated rationale
This PR changes local host probes from localhost to 127.0.0.1 in bin/lib/nim.js, but the accompanying tests for that module were not updated and still assert http://localhost:... in test/nim.test.js. More importantly, this means we are now applying the IPv4-only change to NIM health checks too, even though the PR description is specifically about local inference detection for Ollama/vLLM onboarding.

Please either:
- update test/nim.test.js and clarify in the PR description that NIM health/status probes are intentionally being changed too, or
- revert the bin/lib/nim.js changes and keep this PR scoped to the onboarding/local-inference paths only.
As-is, the patch looks partially applied/internally inconsistent.

chandler-barlow · 2026-04-05T22:03:09Z

addressing the comments here

wscurran added bug Something fails against expected or documented behavior NemoClaw CLI labels Mar 19, 2026

wscurran added Getting Started labels Mar 19, 2026

fix(onboard): use 127.0.0.1 instead of localhost for local inference …

abc1036

…detection

dumko2001 force-pushed the fix/h6-vllm-localhost-bind branch from aed6e2a to abc1036 Compare March 21, 2026 21:33

coderabbitai Bot reviewed Mar 21, 2026

View reviewed changes

fix(onboard): align local inference health checks to 127.0.0.1 and fi…

9d2653c

…x tests

wscurran assigned kjw3 Mar 23, 2026

chandler-barlow self-assigned this Mar 25, 2026

mafueee pushed a commit to mafueee/NemoClaw that referenced this pull request Mar 28, 2026

fix(sandbox): fix create ordering race, dual-registry credentials, an…

ffeaf0d

…d policy identity clearing (NVIDIA#176)

brandonpelfrey reviewed Mar 31, 2026

View reviewed changes

chandler-barlow mentioned this pull request Apr 5, 2026

fix(onboard): use 127.0.0.1 instead of localhost for local inference detection #1518

Closed

3 tasks

chandler-barlow closed this Apr 5, 2026

chandler-barlow mentioned this pull request Apr 9, 2026

fix(onboard): use 127.0.0.1 instead of localhost for local inference … #1716

Closed

3 tasks

cv mentioned this pull request May 17, 2026

chore(security): triage code scanning backlog by risk/reward #3654

Open

6 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix(onboard): use 127.0.0.1 instead of localhost for local inference detection#176

fix(onboard): use 127.0.0.1 instead of localhost for local inference detection#176
dumko2001 wants to merge 2 commits into
NVIDIA:mainfrom
dumko2001:fix/h6-vllm-localhost-bind

dumko2001 commented Mar 17, 2026 •

edited by coderabbitai Bot

Loading

Uh oh!

wscurran commented Mar 19, 2026

Uh oh!

cv commented Mar 21, 2026

Uh oh!

coderabbitai Bot commented Mar 21, 2026 •

edited

Loading

Walkthrough

Changes

Estimated code review effort

Poem

❌ Failed checks (1 warning)

Uh oh!

coderabbitai Bot left a comment

Uh oh!

brandonpelfrey left a comment

Uh oh!

chandler-barlow commented Apr 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Conversation

dumko2001 commented Mar 17, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Rationale

Changes

Verification Results

Leading Standards

Summary by CodeRabbit

Uh oh!

wscurran commented Mar 19, 2026

Uh oh!

cv commented Mar 21, 2026

Uh oh!

coderabbitai Bot commented Mar 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Poem

❌ Failed checks (1 warning)

Uh oh!

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

brandonpelfrey left a comment

Choose a reason for hiding this comment

Uh oh!

chandler-barlow commented Apr 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

dumko2001 commented Mar 17, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented Mar 21, 2026 •

edited

Loading