The tunnel-docker job in v1.7.3 release failed with:
error: failed to unpack package `serde_json v1.0.149`
Caused by: failed to open `/usr/local/cargo/registry/src/.../serde_json-1.0.149/.cargo-ok`
Caused by: File exists (os error 17)
Root cause: BuildKit's default cache-mount sharing is "shared" — both
linux/amd64 and linux/arm64 build stages mount the SAME on-disk cache
dir. Cargo's registry source extraction is non-atomic; both arches
race on `tar -xzf serde_json-1.0.149.crate` into the same destination,
and the loser hits EEXIST mid-unpack.
Fix: scope each cache mount with `id=cargo-registry-${TARGETPLATFORM}`
(and matching for cargo-git + target). BuildKit then keeps separate
on-disk caches per architecture — no race. Per-arch warm-build speedup
is preserved (each cache fills with that arch's pre-built deps); the
only loss is one cache miss per arch on the first build after this
change, which we already paid in v1.7.3.
The target/ mount is also platform-scoped since target/ holds compiled
object files for a single ABI; sharing across arches would either miss
or, worse, link wrong-ABI objects together.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tunnel Node
HTTP tunnel bridge server for MasterHttpRelayVPN "full" mode. Bridges HTTP tunnel requests (from Apps Script) to real TCP connections.
Architecture
Phone → mhrv-rs → [domain-fronted TLS] → Apps Script → [HTTP] → Tunnel Node → [real TCP] → Internet
The tunnel node manages persistent TCP and UDP sessions. TCP sessions are real TCP connections to a destination server; UDP sessions are connected UDP sockets to one destination host:port. Data flows through a JSON protocol:
- connect — open TCP to host:port, return session ID
- data — write client data, return server response
- udp_open — open UDP to host:port, optionally send the first datagram
- udp_data — send one UDP datagram, or poll for returned datagrams when
dis omitted - close — tear down session
- batch — process multiple ops in one HTTP request (reduces round trips)
Deployment
Cloud Run
cd tunnel-node
gcloud run deploy tunnel-node \
--source . \
--region us-central1 \
--allow-unauthenticated \
--set-env-vars TUNNEL_AUTH_KEY=$(openssl rand -hex 24) \
--memory 256Mi \
--cpu 1 \
--max-instances 1
Docker — prebuilt image (any VPS)
The fastest path. Pull a prebuilt image and run it; no Rust toolchain needed on the VPS.
# Generate a strong secret. Save it — you'll paste the same value into CodeFull.gs.
SECRET=$(openssl rand -hex 24)
echo "Your TUNNEL_AUTH_KEY: $SECRET"
# Pull + run.
docker run -d \
--name mhrv-tunnel \
--restart unless-stopped \
-p 8080:8080 \
-e TUNNEL_AUTH_KEY="$SECRET" \
ghcr.io/therealaleph/mhrv-tunnel-node:latest
The :latest tag tracks the most recent release. To pin a specific version (recommended for production), use ghcr.io/therealaleph/mhrv-tunnel-node:v1.5.0 (or whatever release you're on). Image is available for linux/amd64 and linux/arm64.
docker-compose.yml if you prefer:
services:
tunnel:
image: ghcr.io/therealaleph/mhrv-tunnel-node:latest
restart: unless-stopped
ports:
- "8080:8080"
environment:
TUNNEL_AUTH_KEY: ${TUNNEL_AUTH_KEY}
Then TUNNEL_AUTH_KEY=your-secret docker compose up -d.
Docker — build from source
If you'd rather build the image yourself (or add custom changes):
cd tunnel-node
docker build -t tunnel-node .
docker run -p 8080:8080 -e TUNNEL_AUTH_KEY=your-secret tunnel-node
Direct binary
cd tunnel-node
cargo build --release
TUNNEL_AUTH_KEY=your-secret PORT=8080 ./target/release/tunnel-node
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
TUNNEL_AUTH_KEY |
Yes | changeme |
Shared secret — must match TUNNEL_AUTH_KEY in CodeFull.gs |
PORT |
No | 8080 |
Listen port (Cloud Run sets this automatically) |
Protocol
Single op: POST /tunnel
{"k":"auth","op":"connect","host":"example.com","port":443}
{"k":"auth","op":"data","sid":"uuid","data":"base64"}
{"k":"auth","op":"close","sid":"uuid"}
Batch: POST /tunnel/batch
{
"k": "auth",
"ops": [
{"op":"data","sid":"uuid1","d":"base64"},
{"op":"udp_data","sid":"uuid2","d":"base64"},
{"op":"close","sid":"uuid3"}
]
}
→ {"r": [{...}, {...}, {...}]}
Health check: GET /health → ok
Performance: deployment count and pipeline depth
The mhrv-rs client runs a pipelined batch multiplexer in full mode. Each Apps Script round-trip takes ~2s, so the client fires multiple batch requests concurrently — the pipeline depth equals the number of configured script deployment IDs (minimum 2, no upper cap).
More deployments = more concurrent batches hitting the tunnel-node = lower per-session latency. With 6 deployments, a new batch arrives every ~0.3s instead of every 2s.
The tunnel-node itself is stateless per-request (sessions are keyed by UUID), so it handles concurrent batches naturally. For best results, deploy 3–12 Apps Script instances across separate Google accounts and list all their deployment IDs in the client config.