Files
redapricot/docs/architecture.md
T
2026-08-15 17:31:35 +08:00

292 lines
17 KiB
Markdown

# redapricot architecture
This document explains *how* redapricot is built and *why*. For the exact bytes
on the wire, read [PROTOCOL.md](../PROTOCOL.md).
## 1. Roles and topology
```
┌───────────────────────── public internet ─────────────────────────┐
│ │
┌──────────┐ MC handshake (Intent 2/…) ┌───────────────┐ │
│ Player │ ───────────────────────────────▶│ │ │
└──────────┘ raw Minecraft bytes │ Hub │ │
│ (Java/Vert.x)│ │
┌──────────┐ Intent 17, magic 0x01 │ │ │
│ Client │ ◀──────── control session ──────│ • pattern reg │ │
│ (Go) │ ────────────────────────────────│ • CID table │ │
│ │ Intent 17, magic 0x02 │ • mux demux │ │
│ │ ═════════ worker conns ═════════│ │ │
└──────────┘ multiplexed player streams └───────────────┘ │
│ │
▼ MC bytes (+ optional HAProxy v2) │
┌───────────────┐ │
│ Real MC server│ (behind NAT, next to the client) │
└───────────────┘ │
```
Everything reaches the hub on **one TCP port**. The hub distinguishes three
kinds of inbound connection purely from the first Minecraft **Handshake**:
| Handshake `Intent` | Handled as |
|--------------------|------------|
| `17` + magic `0x01` | a **control session** from a client |
| `17` + magic `0x02` | a **worker connection** from a client |
| `18` | reserved (management/status) — never treated as a player |
| anything else | a **player** to be pattern-matched and tunneled |
Because players use ordinary intents (`1` status, `2` login, `3` transfer),
**vanilla clients need no changes**.
## 2. Connection lifecycle
### 2.1 Client establishes a control session
```
Client Hub
│ TCP connect │
│─ Handshake(Intent=17, addr=hex(SHA3-224(PSK))) ─▶ verify addr == expected
│ │
│ (both derive Phase-A keys = ChaCha20(SHA3-256(PSK ‖ dir)))
│─ Frame#1 [magic=0x01, rand, ts] ──────▶ check |now-ts| ≤ window
│ (both switch to Phase-B keys = ChaCha20(SHA3-256(rand‖ts ‖ dir)))
│◀──────────── Frame [SessionReady] ─────│
│─ Register("mc\.example\.com") ────────▶ patterns["mc\.example\.com"] = (regex, session)
│◀──────────── RegisterAck ──────────────│
│ ... periodic Ping/Pong ... │
```
Only frame #1 is encrypted with the PSK-derived key; a fresh random `rand‖ts`
becomes the per-connection key for everything after, so two connections never
share a keystream beyond that first frame.
### 2.2 A player arrives and is tunneled
```
Player Hub Client Destination
│─ Handshake(addr="mc.example.com", Intent=2)─▶ normalize + regex-match
│ │ pause player socket,
│ │ buffer bytes, mint CID
│ │─ ControlRequest(CID, pattern, ip:port) ─▶
│ │ allocate worker+stream
│ │◀──────── SYN(streamId, CID) ────────────│
│ │ takePending(CID) → bind dial destination,
│ │ forward buffered bytes write HAProxy v2 hdr
│ │─ DATA(streamId, handshake…) ───▶ ── handshake ──▶│
│ resume ─────────────────────────────│ bridge stream ⇄ dest
│══════════════ player bytes ══ DATA ══▶│════ DATA ═══▶ dest.write │
│◀═══════════ dest bytes ═══ DATA ══════│◀═══ DATA ════ dest.read │
│ player closes ──────────────────────│─ FIN(streamId) ────────▶ close dest │
```
Key points:
* **Patterns are regexes.** Each registered pattern is a case-insensitive
regular expression, matched against the *whole* normalized hostname (anchored,
first match wins). The hub echoes the **matched pattern string** — not the
player's hostname — in `ControlRequest`, so the client can look it straight up
in its own route table. Invalid patterns are rejected at registration with a
non-zero `RegisterAck` status.
* The hub **pauses** the player socket the instant it matches, so no player
bytes are lost while the takeover is arranged; the buffered handshake is
forwarded **verbatim**, so the real server sees exactly what the player sent
(including the original hostname — used for virtual-host routing there).
* **CID** is 16 random bytes minted by the hub and delivered only over the
encrypted control session, so only the intended client learns it. Any worker
connection presenting the correct CID is allowed to take over — that secrecy
is what binds a worker stream to the right pending player without any explicit
client identity.
* Disconnects are symmetric: player-close → hub sends `FIN` → client closes the
destination; destination-close → client sends `FIN` → hub closes the player.
## 3. Multiplexing (worker connections)
A worker connection is one encrypted TCP link carrying many **streams**. The
frame is intentionally tiny (PROTOCOL.md §7):
```
[plaintext VarInt length][ FrameType u8 | StreamID VarInt | Data… ] (payload encrypted)
```
Only the client opens streams (`SYN`), so stream-id allocation is a simple
per-connection counter with no coordination.
### 3.1 Pool & allocation
The client keeps 1…`maxConn` worker connections and places each new stream on
the **least-loaded** one. The pool grows **breadth-first**: it dials out to
`maxConn` before stacking streams, so that no single TCP connection ever becomes
the shared point of failure for every player on the tunnel (PROTOCOL.md §7.1):
```
pick least-loaded conn; use it
if leastLoaded.streams >= 1 and pool.size + dialsInFlight < maxConn:
dial another worker conn in the background # the stream just placed does not wait
```
The new connection becomes the least-loaded one and picks up subsequent streams.
Once the pool is at `maxConn`, streams stack on the least-loaded connection;
going past 8 active streams there is logged as pool saturation but is not an
error.
Two properties of the dialing path matter as much as the placement rule:
* A dial is **never performed while holding the pool lock** — session
establishment is network I/O, and one unresponsive hub must not park every
other player behind it.
* Only when the pool is *empty* does a caller dial synchronously, and then
exactly one caller dials while the others wait on its result, so a burst of
arrivals cannot open a burst of redundant connections.
The e2e test `TestConcurrentStreamsUseMultipleConns` drives 20 simultaneous
streams with `maxConn=4` and confirms they spread over more than one connection
without exceeding the cap; `TestAllocateDoesNotWedgePoolOnStalledHub` covers the
stalled-dial path.
## 4. Encryption
* **Cipher:** ChaCha20 (RFC 8439) as a raw stream cipher over frame *payloads*.
The length prefix is plaintext, which makes the cipher **phase switch** at
rekey trivial (a reader always knows exactly how many ciphertext bytes belong
to the current frame and never decrypts the next frame with the wrong key).
* **Keys:** `SHA3-256(phaseKey ‖ 0x01)` for client→server and
`SHA3-256(phaseKey ‖ 0x02)` for server→client. Distinct per-direction keys
with a fixed zero nonce avoid a two-time pad without nonce management.
* **Interop:** Java's JCE `ChaCha20` and Go's `x/crypto/chacha20` produce byte-
identical keystreams (including across partial-block, arbitrarily-split
writes), and both `crypto/sha3` implementations agree — verified directly and
pinned by unit tests on both sides against a shared SHA3-224 vector.
## 5. Threading model
* **Hub:** a single Vert.x verticle instance. All accepted connections are
handled on that verticle's one event loop, so the pattern registry, CID table,
and per-connection state are touched by a single thread — no locks on the hot
path (concurrent maps are used only defensively). Every socket operation is
non-blocking; crypto is CPU-cheap. This trades multi-core scaling for
simplicity and correctness.
* **Client:** goroutine-per-concern. One goroutine reads each connection
(control or worker); `WriteFrame` is mutex-serialized so many stream goroutines
can share a worker connection safely. Each stream has two goroutines: `run`
pumps destination → hub, and `writeLoop` is the only writer to the
destination, draining a per-stream queue fed by the worker readLoop. The
readLoop itself never writes to a destination, so a stalled destination can
never block frame dispatch for other streams.
## 6. Back-pressure & flow control
Three mechanisms operate at different granularities:
* **Per-stream credit windows** (PROTOCOL.md §7.3; the windows are exchanged
at session establishment): each stream direction has an independent byte
budget equal to the receiver's advertised window (default 256 KiB). A sender
that exhausts a
stream's window pauses *only that stream's source* — the hub pauses the one
player socket, the client parks the one destination-reader goroutine. Credit
is granted back (`WND` frames, batched at half-window) as bytes are actually
written to the terminal socket. The result: a slow player or slow destination
jams its own stream at a bounded buffer size and nothing else. This is what
eliminates head-of-line blocking between streams.
* **Aggregate TCP back-pressure** on each worker connection: when the shared
socket itself is congested (total bandwidth, not one stream), the hub parks
all sending players until it drains, and the client's `WriteFrame` blocks.
This is fair — when the pipe is genuinely full, everyone should slow down.
* **Client egress shaping** (optional, `maxBandwidth`; `client/shaper.go`): a
rate cap on everything the client sends to the hub, across all worker conns.
The first two mechanisms have no time dimension. A credit window bounds how many
bytes are *in flight*, and TCP back-pressure only reacts once the pipe is already
full — which on a residential uplink is too late. One player loading chunks fills
the line, the standing queue grows to seconds, and every other player's keepalive
times out. Nothing in §7.3 prevents that: each stream is individually
well-behaved, and collectively they still overrun the link.
The shaper closes that gap with a token bucket for the rate and start-time fair
queueing for the split. A global virtual clock advances with each grant; every
stream remembers where its last request finished, and a new request is stamped
`max(stream.vfinish, vclock)`. Lowest stamp wins. A stream that keeps sending
pushes its own stamp further out and yields; a stream returning from idle is
clamped back to the clock, so it cannot bank credit for time it did not use, but
is not penalised for the idleness either. A stream sending a few hundred bytes
gets a nearer stamp than one sending a full chunk, so keepalives and chat overtake
bulk terrain data for free. One stream alone still gets the entire rate.
Two details keep bursts cheap. The bucket banks 200 ms of transmission, so a
player joining spends it at once instead of paying for the cap in visible
chunk-loading latency. And the DATA chunk shrinks to ~20 ms of transmission when
the rate is low (floor 4 KiB), because a fixed 32 KiB chunk is a 256 ms slot at
1 Mbps — long enough dead air to drag the other players towards the very timeout
the cap exists to prevent. Above ~13 Mbps the chunk stays at the usual 32 KiB.
This is entirely client-local: nothing about it appears on the wire, and the hub
is unaware. Only DATA is shaped — delaying a `FIN`, `WND` or `PONG` would cause
the false-death detection §7.4 exists to avoid.
The window also bounds memory: a stream can hold at most one window of
undelivered data per direction (the client's pre-connect handshake buffer is
covered by the same bound). With stream resumption enabled (§7) the *sender*
holds a second window — the bytes it has sent but the peer has not yet credited,
kept so they can be retransmitted after an outage. That is not a new bound so
much as the existing one made symmetric: the region is exactly what flow control
already declared outstanding, which is why resumption needs no cap of its own.
Per-stream flow control is mandatory: the hub rejects a session whose Rekey
lacks the STREAM_FC flag, and the client rejects a hub that does not echo it —
peers that predate the mechanism cannot connect at all.
What remains (by design) is TCP-level head-of-line blocking: a lost packet on
a worker connection stalls all its streams for one retransmit. That is inherent
to mux-over-TCP; the connection pool is the mitigation, and a datagram
transport (QUIC) would be the escape hatch if it ever matters.
## 7. Failure & recovery
* **Control session drop:** the client retries immediately, then backs off to a
10s cap, and re-registers all patterns. Existing worker connections and their
live streams are unaffected — they ride worker conns, which a control-session
close never touches. The hub meanwhile keeps that session's routes as
*orphaned* for `registrationGraceMs` (PROTOCOL.md §5.2) and **holds** players
arriving on them instead of refusing them, replaying the control request once
a client re-registers the pattern. Without that, the reconnect window is one
in which every new player is told there is no such server.
* **Worker connection drop:** the connection leaves the pool either way. What
happens to its streams depends on whether STREAM_RESUME was negotiated:
* *without it* — every stream is torn down (destinations closed) and the hub
closes the corresponding player sockets, as it always did;
* *with it* — the streams are **hung** instead (PROTOCOL.md §7.5). The
destination sockets stay open, the hub pauses the player sockets and holds
them for its grace period, and the client reattaches each stream over a
freshly dialed connection, replaying byte-exactly from the offset the peer
reports. Players see a stall rather than a disconnect.
* **Pending timeout:** if no worker takes over a matched player within
`pendingTimeoutMs`, the hub drops the pending entry and closes the player.
* **Bad PSK / bad timestamp / bad magic:** the hub closes the TCP connection;
the client's session establishment fails fast.
Resumption is worth the machinery because a worker connection is only the
*middle* leg of every stream it carries. When it dies both terminal sockets are
usually still healthy, so the old behaviour discarded working connections
because a replaceable transport failed — one conntrack expiry disconnected every
player sharing that connection. It also has to be byte-exact rather than
best-effort: bytes handed to a dying socket are lost with no notification and the
frame cipher cannot be resynchronized, so an approximate reattach would splice
the tunneled protocol mid-packet, which is worse than a clean close.
Notably, resumption does not depend on the control session. A blip usually kills
both, and the reattach path needs only a worker connection, so recovery does not
wait on the control reconnect backoff.
## 8. Known limitations
1. No AEAD — payload integrity/authenticity is not cryptographically guaranteed.
2. TCP-level head-of-line blocking within a worker connection (lost packets;
see §6) — per-stream flow control removes the application-level variant only.
3. Single-event-loop hub (see §5) bounds throughput to one core.
4. `Intent 18` is reserved but only stubbed (the hub logs and closes).
5. Pattern ownership is last-writer-wins; two clients registering the identical
pattern string will silently reassign it. Overlapping-but-distinct regexes are
both kept, and when several match one hostname the winner is unspecified.
These are deliberate scope choices for a connectivity-focused P2P tool, not
oversights; each is a small, well-isolated change away from being hardened.