Issue Title: Bug: Codebase indexing fails with "fetch failed" when using Gemini embedder

[GreyHorizonX]

What specific problem does this solve?

Description
When configuring the Codebase Indexing feature with Gemini as the embedding provider, the initial scan fails with a fetch failed error. This prevents the user from creating a semantic index of their project, blocking a core feature.

Steps to Reproduce
Navigate to the ”Codebase Indexing“ settings within the extension.
Enable the feature by checking “Enable Codebase Indexing”.
Set the "Embedder Provider“ to Gemini.
Enter a valid Gemini API Key.
Keep the default model: text-embedding-004.
Set the "Qdrant URL" to a running local instance (e.g., http://localhost:6333).
Click the "Start” button to initiate indexing.
The status immediately shows the “Error - Failed during initial scan: fetch failed”.
Expected Behavior
The indexing process should start successfully. The extension should connect to the Gemini API to generate embeddings for the codebase and then store them in the specified local Qdrant instance.

Actual Behavior
The process fails immediately with a generic fetch failed network error.

and the log：

[cloud-settings] Error fetching organization settings:
Error: fetch failed
TypeError: fetch failed
    at node:internal/deps/undici/undici:13392:13
    at processTicksAndRejections (node:internal/process/task_queues:95:5)
    at globalThis.fetch (file:///root/.vscode-server/bin/e249dada235c2083c83813bd65b7f4707fb97b76/out/vs/workbench/api/node/extensionHostProcess.js:174:20102)
    at nae.fetchSettings (/root/.vscode-server/extensions/packages/cloud/src/CloudSettingsService.ts:84:21)
    at aV.callback (/root/.vscode-server/extensions/packages/cloud/src/CloudSettingsService.ts:43:12)
    at aV.executeCallback (/root/.vscode-server/extensions/packages/cloud/src/RefreshTimer.ts:146:19)

so i find the

packages/cloud/src/CloudSettingsService.ts:84:21
try {
			const response = await fetch(`${getRooCodeApiUrl()}/api/organization-settings`, {
				headers: {
					Authorization: `Bearer ${token}`,
				},
			})

			if (!response.ok) {
				this.log(
					"[cloud-settings] Failed to fetch organization settings:",
					response.status,
					response.statusText,
				)
				return false
			}

packages/cloud/src/config.ts
export const PRODUCTION_CLERK_BASE_URL = "https://clerk.roocode.com"
export const PRODUCTION_ROO_CODE_API_URL = "https://app.roocode.com"

export const getClerkBaseUrl = () => process.env.CLERK_BASE_URL || PRODUCTION_CLERK_BASE_URL
export const getRooCodeApiUrl = () => process.env.ROO_CODE_API_URL || PRODUCTION_ROO_CODE_API_URL

I executed and It can connect, so I really don't know what's wrong.

curl -v "https://app.roocode.com/api/organization-settings"

To prove my network environment, I manually test the Gemini API:

I don't know if there are issues with the steps above. This problem has troubled me for a long time; it's always been fetch failed. I'm looking forward to your good news.

Additional context (optional)

No response

Roo Code Task Links (Optional)

No response

Request checklist

• I've searched existing Issues and Discussions for duplicates
• This describes a specific problem with clear impact and context

Interested in implementing this?

• Yes, I'd like to help implement this feature

Implementation requirements

• I understand this needs approval before implementation begins

How should this be solved? (REQUIRED if contributing, optional otherwise)

No response

How will we know it works? (Acceptance Criteria - REQUIRED if contributing, optional otherwise)

No response

Technical considerations (REQUIRED if contributing, optional otherwise)

No response

Trade-offs and risks (REQUIRED if contributing, optional otherwise)

No response

[roomote]

Hi @teranexa! Im here to help out the maintainers and am going to see if I can fix this issue or at least get close enough to be helpful!

I can see the issue is related to a "fetch failed" error when using Gemini embedder for codebase indexing. The error appears to be coming from the CloudSettingsService when trying to fetch organization settings. Ill investigate this and work on a fix.

Thanks for the detailed bug report with screenshots and your debugging attempts - theyre very helpful!

[roomote]

Hi @teranexa! Ive created a pull request to address the "fetch failed" error you were experiencing: #6627

The PR adds:

1. Retry mechanism with exponential backoff - The fetch operation will now automatically retry up to 3 times with increasing delays (1s, 2s, 4s) to handle transient network issues
2. Enhanced error logging - When a fetch fails, it now logs detailed diagnostic information including:
   • Error name, message, and stack trace
   • Proxy configuration (HTTP_PROXY, HTTPS_PROXY, NO_PROXY)
   • Node.js and VSCode versions
   • Helpful suggestions for common causes
3. Network diagnostics - The error handler now performs network diagnostics to help identify the root cause

This should provide much more helpful information when debugging network-related issues with the Gemini embedder or any other features that rely on CloudSettingsService.

The CI checks are currently running on the PR. Once they pass and the PR is merged, you should see much more detailed error messages that will help identify the root cause of the fetch failures.

[GreyHorizonX]

Thank you !

[Juno933]

@teranexa Did you running Qdrant instance? I had the same problem, but with cloud instance. I put wrong url. When i fixed to correct one, error dissapeared.

[GreyHorizonX]

@SoFo12 I tried two situations; however, Qdrant uses a local Docker setup, which still has certain problems. Only in the OLLMA+embedding model, in a local quadrant situation, can I successfully use it.