Restructure Troubleshooting into a hub with per-area pages (#814)

* Restructure Troubleshooting into a hub with per-area pages

- Add a Troubleshooting hub (/help/troubleshooting) with icon/chip cards and a "Still stuck?" CTA
- Split NetBird Client troubleshooting into an overview + per-OS pages (Linux, Windows, macOS, Android, iOS)
- Split Self-hosted troubleshooting into an overview + per-area pages (installation, IdP, dashboard, certificates, connectivity, database)
- Split "Report bugs and issues" into Community Support and NetBird Support pages
- Add Troubleshooting resource connectivity and a NetBird Cloud pending-approval page
- Add DNS troubleshooting Issue 8 (Windows NRPT rule blocked by a lingering GPO)
- Cross-reference the new pages from networks, DNS, and reverse-proxy docs; update nav

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* Address review: client terminology, dead props, labels, cross-links

- Use "client" instead of "agent" across the client troubleshooting pages (headings, prose, anchors)
- Remove unused source: props from the Troubleshooting hub tiles
- Relabel the "NetBird Cloud" grouping to "Cloud & identity" (SSO/provisioning also apply to self-hosted)
- Add a Tiles title on the report-bug landing; add reverse-proxy -> resource-connectivity cross-link
- Fix comma splices introduced by the em-dash cleanup in relayed-connections

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* Add client-side hash redirect for moved self-hosted anchors

Old deep links like /selfhosted/troubleshooting#debugging-turn-connections now
forward to the per-area page, since next.config redirects can't act on the URL fragment.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* Apply docs-skill review: conventions + reshape area pages

- "open source" (no hyphen), expand NRPT on first use, descriptive alt text + captions on TURN images
- Fix inherited "Netbird" casing in the client glossary
- Reshape the six self-hosted area pages to Symptom -> likely causes (ordered) -> Fix -> Confirm, preserving anchored headings

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* Fix two typos in client glossary (CodeRabbit)

- "nunning" -> "running" in the glossary
- possessive "it's" -> "its" in the routing-table sentence

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* docs: fix two broken links in troubleshooting pages

- database: point the "upgrade path" link at /selfhosted/maintenance/upgrade;
  selfhosted-quickstart has no #upgrade anchor so the old link landed at page top
- client: add HashRedirect so old #net-bird-agent-status deep links forward to
  the renamed #net-bird-client-status section on the same page

* docs: address review follow-ups (deep-link redirects + client casing)

- self-hosted troubleshooting: extend the HashRedirect map with the per-issue
  (###-level) anchors from the old single page, so old deep links land on the
  exact sub-section of the new area page rather than just the page top
- client glossary: lowercase "NetBird client" in the peer-a/peer-b entries
  (house convention) and fix "linux" -> "Linux"

* docs: review polish — fix image class + first-use acronym glosses

- connectivity: fix bad CSS class imagewrapper-nig -> imagewrapper on the
  TURN-test screenshot (the typo'd class matched no style and broke zoom)
- gloss acronyms on first use: GPO (DNS Issue 8), IdP/SSO (identity-provider),
  ACME (certificates), CORS (dashboard)

---------

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Jack Carter <128555021+SunsetDrifter@users.noreply.github.com>

src/pages/selfhosted/troubleshooting/dashboard.mdx

@@ -0,0 +1,30 @@
+export const description =
+  "Self-hosted NetBird dashboard troubleshooting: blank pages and unauthorized or 403 errors."
+
+# Dashboard issues
+
+Problems with the self-hosted dashboard. For other areas, start from [Troubleshooting](/selfhosted/troubleshooting).
+
+## Dashboard shows blank page
+
+**Symptom**: The dashboard loads but renders a blank page, sometimes with errors in the browser console.
+
+**Likely causes and fixes** (most common first):
+
+1. **The dashboard can't reach the Management API.** Open the browser console (F12 → Console) and look for failed requests. Confirm the dashboard's configured API URL is correct and reachable from the browser.
+2. **CORS (Cross-Origin Resource Sharing) is blocking the API behind a custom reverse proxy.** Serve the API on the same origin as the dashboard, or set the correct CORS headers on your proxy.
+3. **A stale cached bundle.** Hard-reload the page or clear the browser cache.
+
+**Confirm**: Reload the dashboard. The login or peers view renders, and the console shows no failed API requests.
+
+## "Unauthorized" or "403" errors
+
+**Symptom**: API calls return 401 Unauthorized or 403 Forbidden.
+
+**Likely causes and fixes** (most common first):
+
+1. **An expired or invalid token.** Re-authenticate. For direct API access, use a valid Personal Access Token (PAT).
+2. **The user lacks permission** for the action. Check the user's role in the dashboard.
+3. **Management can't validate the token.** Review `docker compose logs management` for token-validation errors.
+
+**Confirm**: Re-run the action. It succeeds, and the Management logs show no auth errors.