Skip to content

[Brev][CLI&UX] connect logs success and failure for same port forward #3334

@zNeill

Description

@zNeill

Description

Description
On every nemoclaw connect invocation on the Brev VM, the CLI emits a success line "✓ Forwarding port 18789 to sandbox brev-cmd in the background" and, in the same block, a contradictory failure line "Failed to re-establish the dashboard port forward. Run openshell forward start --background PORT brev-cmd manually." Connect itself succeeds and the in-sandbox shell + inference work, so this is a logging / UX bug rather than a functional regression — but the output is misleading because the user is told the forward both worked AND failed in adjacent lines.

Separately, every connect invocation also prints "Dashboard port forward to 'brev-cmd' is missing or dead. Re-establishing..." at the very top even immediately after a previous successful connect, which suggests the forward-state probe does not recognise an existing healthy forward.Environment

Device:        Brev command line
OS:            Ubuntu 22.04.5 LTS
Architecture:  x86_64
Node.js:       v22.22.2
npm:           10.9.7
Docker:        29.1.3 (build f52814d)
OpenShell CLI: openshell 0.0.36
NemoClaw:      v0.0.38
OpenClaw:      2026.4.24
Steps to Reproduce
1. Fresh-install NemoClaw on a Brev Ubuntu 22.04 VM:
     curl -fsSL https://www.nvidia.com/nemoclaw.sh | bash
2. Complete onboard with NVIDIA Endpoints, default model, sandbox name "brev-cmd"
3. Run: nemoclaw brev-cmd connect
4. Observe the CLI output before the sandbox shell prompt.
5. Repeat step 3 — the same contradictory message block appears every time.
Expected Result

Either a single accurate status line ("✓ Forwarding port 18789 ..." when the forward actually worked, and nothing else), or one clear failure message when it didn't. The top-line "missing or dead. Re-establishing..." should only appear when the forward is actually missing, not on every connect.Actual Result

A success and a failure message appear in the same block, in that order. Connect proceeds successfully despite the "Failed to re-establish..." line, and the inference round-trip works. Verbatim CLI output:

$ nemoclaw brev-cmd connect

Dashboard port forward to 'brev-cmd' is missing or dead.
Re-establishing...
→ Found forward on sandbox 'brev-cmd'
✓ Stopped forward of port 18789 for sandbox brev-cmd
✓ Forwarding port 18789 to sandbox brev-cmd in the background
Access at: http://127.0.0.1:18789/
Stop with: openshell forward stop 18789 brev-cmd
Failed to re-establish the dashboard port forward.
Run openshell forward start --background PORT brev-cmd manually.

✓ Connecting to sandbox 'brev-cmd'
Inside the sandbox, run openclaw tui to start chatting with the agent.
Type /exit to leave the chat, then exit to return to the host shell.

Logs

Reproduced twice in the same session on the Brev VM. Both runs produced the identical sequence above. Found while executing DevTest task T6003300 (Phase 1, brev-command folder, v0.0.38 manual). Host-side capture saved at /tmp/connect-repro.log.

Bug Details

Field Value
Priority Unprioritized
Action Dev - Open - To fix
Disposition Open issue
Module Machine Learning - NemoClaw
Keyword NemoClaw, NemoClaw_CLI&UX, NEMOCLAW_GH_SYNC_APPROVAL

[NVB#6163982]

Metadata

Metadata

Assignees

Labels

NV QABugs found by the NVIDIA QA Teamarea: cliCommand line interface, flags, terminal UX, or outputplatform: brevAffects Brev hosted development environments

Type

No fields configured for Bug.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions