fix(server): recover a deleted tunnel when Cloudflare says "Tunnel not found"

[juliusmarminge]

When the relay deletes an idle tunnel, the host's connector is supposed to notice the rejection and ask for a replacement at the same hostname. It only recognized three exact messages, and Cloudflare now answers a deleted tunnel with Unauthorized: Tunnel not found. A user's host retried that dead tunnel 503 times instead of recovering. Turning on relay cleanup on 2026-10-06 made this path common, so cleanup is back in dry-run until this ships.

• Any Unauthorized: registration error now counts as a rejection, alongside the unprefixed messages seen so far. Recovery still waits for four rejections in a row, which absorbs the brief rejections while a new tunnel's token propagates.
• A connector that has not registered a connection three minutes after starting requests recovery anyway, and again every three minutes until it connects. When the tunnel is still live the relay hands back the same tunnel, so the connector is left running. This covers the next wording change and any other way a tunnel dies silently.

Tests: the reported line is now matched, and a new TestClock test covers the fallback. Reverting either change fails its test.

Built with Claude Opus 5.5 in Claude Code.

🤖 Generated with Claude Code
Fixes #16399

---

[macroscopeapp]

Approvability

Verdict: Approved at a894e20

Macroscope's review found this PR approvable — This is a contained Cloudflare tunnel recovery fix: it recognizes the newer rejection message and adds a scoped, three-minute fallback for connectors that never register. Existing connected behavior remains unchanged, and the fallback plus message matching are covered by focused tests.

^{You can add or adjust custom eligibility rules. Learn more.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	20.9 KiB	20.9 KiB	0 B (0.0%)	29.3 KiB	✅
Codex	Live turn messages	2	2	0 (0.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.2 KiB	21.2 KiB	0 B (0.0%)	29.3 KiB	✅
Claude	Live turn messages	2	2	0 (0.0%)	8	✅

Baseline: 517188b · PR result: a894e20 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Path: .coderabbit.config.ts
• Review profile: CHILL
• Plan: Team
• Run ID: aed96348-95d4-405c-b0b4-7800be43a8b3

📥 Commits

Reviewing files that changed from the base of the PR and between 517188b and a894e20.

📒 Files selected for processing (2)

• apps/server/src/cloud/ManagedEndpointRuntime.test.ts
• apps/server/src/cloud/ManagedEndpointRuntime.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 5 remain after this review.

---

📝 Walkthrough

Walkthrough

The runtime now recognizes additional tunnel-registration rejection messages. It also monitors new connectors and queues recovery every three minutes until a connection is registered.

Changes

Managed endpoint recovery

Layer / File(s)	Summary
Tunnel rejection matching apps/server/src/cloud/ManagedEndpointRuntime.ts, apps/server/src/cloud/ManagedEndpointRuntime.test.ts	The matcher recognizes additional errors in tunnel-registration output, including Tunnel not found. Tests confirm that the same message in Failed to serve tunnel connection output does not match.
Timed registration recovery apps/server/src/cloud/ManagedEndpointRuntime.ts, apps/server/src/cloud/ManagedEndpointRuntime.test.ts	Each connector starts with an unconnected state. The runtime queues recovery every three minutes until the connector registers a connection. Tests cover repeated recovery for an unconnected connector and no recovery after connection registration.

Priority: ⬆️ High

Estimated code review effort: 2 (Simple) | ~12 minutes

Change: Bug fix

Sequence Diagram(s)

sequenceDiagram
  participant ManagedEndpointRuntime
  participant RegistrationWatcher
  participant ActiveConnector
  participant RecoveryQueue
  ManagedEndpointRuntime->>RegistrationWatcher: Start watcher for connector
  RegistrationWatcher->>ActiveConnector: Check connection status after three minutes
  ActiveConnector-->>RegistrationWatcher: Return connected status
  RegistrationWatcher->>RecoveryQueue: Queue recovery if unconnected

Loading

Suggested reviewers: t3dotgg

Merge Risk: ⚪ Minimal · up to a894e

No actionable merge-blocking issue was established. The recovery change is mergeable after normal checks.

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title clearly identifies the server fix for recovery when Cloudflare reports that a tunnel was not found.
Description check	✅ Passed	The description explains the problem, the recovery changes, and the related tests. It does not use the template headings or clearly state why the change qualifies without prior maintainer approval, an…

✨ Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

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