root fa56134b90 ADR-003 wiring: Bearer token + IP allowlist middleware
Implements the auth posture from ADR-003 (commit 0d18ffa). Two
independent layers — Bearer token (constant-time compare via
crypto/subtle) and IP allowlist (CIDR set) — composed in shared.Run
so every binary inherits the same gate without per-binary wiring.

Together with the bind-gate from commit 6af0520, this mechanically
closes audit risks R-001 + R-007:
  - non-loopback bind without auth.token = startup refuse
  - non-loopback bind WITH auth.token + override env = allowed
  - loopback bind = all gates open (G0 dev unchanged)

internal/shared/auth.go (NEW)
  RequireAuth(cfg AuthConfig) returns chi-compatible middleware.
  Empty Token + empty AllowedIPs → pass-through (G0 dev mode).
  Token-only → 401 Bearer mismatch.
  AllowedIPs-only → 403 source IP not in CIDR set.
  Both → both gates apply.
  /health bypasses both layers (load-balancer / liveness probes
  shouldn't carry tokens).

  CIDR parsing pre-runs at boot; bare IP (no /N) treated as /32 (or
  /128 for IPv6). Invalid entries log warn and drop, fail-loud-but-
  not-fatal so a typo doesn't kill the binary.

  Token comparison: subtle.ConstantTimeCompare on the full
  "Bearer <token>" wire-format string. Length-mismatch returns 0
  (per stdlib spec), so wrong-length tokens reject without timing
  leak. Pre-encoded comparison slice stored in the middleware
  closure — one allocation per request.

  Source-IP extraction prefers net.SplitHostPort fallback to
  RemoteAddr-as-is for httptest compatibility. X-Forwarded-For
  support is a follow-up when a trusted proxy fronts the gateway
  (config knob TBD per ADR-003 §"Future").

internal/shared/server.go
  Run signature: gained AuthConfig parameter (4th arg).
  /health stays mounted on the outer router (public).
  Registered routes go inside chi.Group with RequireAuth applied —
  empty config = transparent group.
  Added requireAuthOnNonLoopback startup check: non-loopback bind
  with empty Token = refuse to start (cites R-001 + R-007 by name).

internal/shared/config.go
  AuthConfig type added with TOML tags. Fields: Token, AllowedIPs.
  Composed into Config under [auth].

cmd/<svc>/main.go × 7 (catalogd, embedd, gateway, ingestd, queryd,
storaged, vectord, mcpd is unaffected — stdio doesn't bind a port)
  Each call site adds cfg.Auth as the 4th arg to shared.Run. No
  other changes — middleware applies via shared.Run uniformly.

internal/shared/auth_test.go (12 test funcs)
  Empty config pass-through, missing-token 401, wrong-token 401,
  correct-token 200, raw-token-without-Bearer-prefix 401, /health
  always public, IP allowlist allow + reject, bare IP /32, both
  layers when both configured, invalid CIDR drop-with-warn, RemoteAddr
  shape extraction. The constant-time comparison is verified by
  inspection (comments in auth.go) plus the existence of the
  passthrough test (length-mismatch case).

Verified:
  go test -count=1 ./internal/shared/  — all green (was 21, now 33 funcs)
  just verify                            — vet + test + 9 smokes 33s
  just proof contract                    — 53/0/1 unchanged

Smokes + proof harness keep working without any token configuration:
default Auth is empty struct → middleware is no-op → existing tests
pass unchanged. To exercise the gate, operators set [auth].token in
lakehouse.toml (or, per the "future" note in the ADR, via env var).

Closes audit findings:
  R-001 HIGH — fully mechanically closed (was: partial via bind gate)
  R-007 MED  — fully mechanically closed (was: design-only ADR-003)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-29 07:11:34 -05:00

187 lines
5.9 KiB
Go

// Package shared provides common HTTP server bootstrap for every
// Lakehouse-Go service. Each cmd/<service> calls Run with its name,
// bind address, and a route-registration callback. The factory wires
// chi, slog, /health, and graceful shutdown identically across all
// five binaries — the place where uniformity beats per-service
// flexibility.
//
// G1+ note: when queryd needs to drain a cgo DuckDB handle on
// shutdown, the simple shared factory will need a per-service hook
// (an io.Closer slice or an OnShutdown callback). For G0 a plain
// chi.Router + http.Server.Shutdown(ctx) is sufficient.
package shared
import (
"context"
"encoding/json"
"errors"
"log/slog"
"net"
"net/http"
"os"
"os/signal"
"syscall"
"time"
"github.com/go-chi/chi/v5"
"github.com/go-chi/chi/v5/middleware"
)
// HealthResponse is the JSON shape returned by /health on every
// service. Service-specific status hooks can extend it post-G0.
type HealthResponse struct {
Status string `json:"status"`
Service string `json:"service"`
}
// RegisterRoutes is the per-service callback that wires its own
// routes onto the shared router AFTER /health has been mounted.
type RegisterRoutes func(r chi.Router)
// Run boots a chi router with slog logging, the /health endpoint,
// and graceful-shutdown handling. Blocks until SIGINT/SIGTERM or a
// fatal listener error.
//
// The logger is constructed locally and used as the request-logging
// sink. Run does NOT mutate the global slog default — callers that
// want their own slog.Default() should set it before calling Run.
// (Per Kimi review #4: shared library functions shouldn't silently
// mutate package globals.)
//
// Three startup gates apply in order:
//
// 1. requireLoopbackOrOverride — refuses non-loopback bind unless
// LH_<SERVICE>_ALLOW_NONLOOPBACK=1 is set. Closes the accidental
// 0.0.0.0 deploy path for R-001.
// 2. requireAuthOnNonLoopback — refuses non-loopback bind when
// auth.token is empty. Mechanically prevents R-001 + R-007's
// worst case: world-reachable bind with no auth layer.
// 3. RequireAuth middleware — runs per-request on registered routes.
// /health stays exempt (mounted on the outer router, before the
// authed group).
//
// Per ADR-003: empty auth.token + empty allowed_ips → middleware is
// a no-op. Smokes and proof harness keep working without setting
// either.
func Run(serviceName, addr string, register RegisterRoutes, auth AuthConfig) error {
if err := requireLoopbackOrOverride(serviceName, addr); err != nil {
return err
}
if err := requireAuthOnNonLoopback(serviceName, addr, auth); err != nil {
return err
}
logger := slog.New(slog.NewJSONHandler(os.Stdout, &slog.HandlerOptions{
Level: slog.LevelInfo,
}))
r := chi.NewRouter()
r.Use(middleware.RequestID)
r.Use(middleware.RealIP)
r.Use(middleware.Recoverer)
r.Use(slogRequest(logger))
// /health stays on the outer router — public, no auth. Operators
// rely on it for liveness probes that don't carry a token.
r.Get("/health", func(w http.ResponseWriter, _ *http.Request) {
w.Header().Set("Content-Type", "application/json")
_ = json.NewEncoder(w).Encode(HealthResponse{
Status: "ok",
Service: serviceName,
})
})
if register != nil {
// Registered routes live inside an auth-gated group so
// RequireAuth applies uniformly without per-handler wiring.
// Empty auth → middleware is a no-op (group is transparent).
r.Group(func(authed chi.Router) {
authed.Use(RequireAuth(auth))
register(authed)
})
}
srv := &http.Server{
Addr: addr,
Handler: r,
ReadHeaderTimeout: 10 * time.Second,
}
ctx, stop := signal.NotifyContext(context.Background(), os.Interrupt, syscall.SIGTERM)
defer stop()
// Race-safe startup: bind the listener synchronously BEFORE
// returning so a fast bind error (e.g. port already in use) is
// surfaced as Run's return value rather than racing the select.
// Per Opus + Qwen BLOCK #1: the prior pattern could drop bind
// errors when ctx.Done already fired or a fast failure happened
// during select setup.
ln, err := newListener(srv.Addr)
if err != nil {
return err
}
errCh := make(chan error, 1)
go func() {
logger.Info("listening", "service", serviceName, "addr", addr)
if err := srv.Serve(ln); err != nil && !errors.Is(err, http.ErrServerClosed) {
errCh <- err
}
close(errCh)
}()
select {
case <-ctx.Done():
logger.Info("shutdown signal received", "service", serviceName)
case err := <-errCh:
if err != nil {
return err
}
// Server exited cleanly without a signal (unlikely but possible).
return nil
}
shutdownCtx, cancel := context.WithTimeout(context.Background(), 15*time.Second)
defer cancel()
if err := srv.Shutdown(shutdownCtx); err != nil {
return err
}
// Drain errCh so a late error from the listener goroutine
// surfaces as the return value instead of leaking. After Shutdown
// the channel will close on graceful exit; if a real error
// landed first we return it.
if err := <-errCh; err != nil {
return err
}
return nil
}
// newListener binds the TCP listener up-front so bind errors are
// returned synchronously to Run's caller. Extracted into its own
// function for testability + to keep Run readable.
func newListener(addr string) (net.Listener, error) {
return net.Listen("tcp", addr)
}
// slogRequest returns a chi middleware that logs each request via slog.
// Replaces chi's default text logger so all log output stays JSON.
func slogRequest(logger *slog.Logger) func(http.Handler) http.Handler {
return func(next http.Handler) http.Handler {
return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
start := time.Now()
ww := middleware.NewWrapResponseWriter(w, r.ProtoMajor)
defer func() {
logger.Info("http",
"method", r.Method,
"path", r.URL.Path,
"status", ww.Status(),
"dur_ms", time.Since(start).Milliseconds(),
"req_id", middleware.GetReqID(r.Context()),
)
}()
next.ServeHTTP(ww, r)
})
}
}