CosmicClash/server/cmd/control-plane/main.go

package main

import (
	"context"
	"database/sql"
	"flag"
	"fmt"
	"net/http"
	"os"
	"os/signal"
	"syscall"
	"time"

	"github.com/cosmic-clash/cosmic-clash/server/api"
	"github.com/cosmic-clash/cosmic-clash/server/domain"
	"github.com/cosmic-clash/cosmic-clash/server/migrations"
	"github.com/cosmic-clash/cosmic-clash/server/observability"
	"github.com/cosmic-clash/cosmic-clash/server/store"
	_ "github.com/jackc/pgx/v5/stdlib"
	"github.com/redis/go-redis/v9"
)

func main() {
	listen := flag.String("listen", ":8080", "HTTP listen address")
	role := flag.String("role", "api", "control-plane role; currently api")
	dsn := flag.String("dsn", os.Getenv("COSMIC_CLASH_POSTGRES_DSN"), "PostgreSQL connection string")
	migrationDir := flag.String("migrations", "migrations", "directory containing numbered SQL migrations")
	redisAddr := flag.String("redis-addr", os.Getenv("COSMIC_CLASH_REDIS_ADDR"), "optional Redis address for the candidate projection")
	redisPrefix := flag.String("redis-prefix", envOrDefault("COSMIC_CLASH_REDIS_PREFIX", "cosmic-clash"), "Redis key prefix")
	redisTTL := flag.Duration("redis-ttl", 60*time.Second, "TTL for transient candidate projection entries")
	workloadSecret := flag.String("workload-secret", os.Getenv("COSMIC_CLASH_WORKLOAD_SECRET"), "HMAC secret for control-plane-issued workload tokens (see workload/signed_token.go); server registration/result submission return 503 until this is set")
	flag.Parse()
	if *role != "api" {
		fatalf("unsupported role %q (only api is implemented)", *role)
	}
	if *dsn == "" {
		fatalf("--dsn or COSMIC_CLASH_POSTGRES_DSN is required")
	}
	if *redisTTL <= 0 {
		fatalf("--redis-ttl must be positive")
	}
	db, err := sql.Open("pgx", *dsn)
	if err != nil {
		fatalf("open PostgreSQL: %v", err)
	}
	defer db.Close()
	startupCtx, startupCancel := context.WithTimeout(context.Background(), 30*time.Second)
	defer startupCancel()
	if err := db.PingContext(startupCtx); err != nil {
		fatalf("ping PostgreSQL: %v", err)
	}
	if err := migrations.Apply(startupCtx, db, *migrationDir); err != nil {
		fatalf("apply migrations: %v", err)
	}
	var candidateIndex api.CandidateIndex
	var redisClient *redis.Client
	if *redisAddr != "" {
		redisClient = redis.NewClient(&redis.Options{Addr: *redisAddr})
		defer redisClient.Close()
		candidateIndex = store.RedisCandidateIndex{Client: redisClient, Prefix: *redisPrefix, TTL: *redisTTL}
	}
	if *workloadSecret == "" {
		fmt.Fprintln(os.Stderr, "control-plane: warning: --workload-secret / COSMIC_CLASH_WORKLOAD_SECRET is unset; server registration and result submission will return 503")
	}
	service := newAPIService(db, *workloadSecret, candidateIndex)
	server := &http.Server{Addr: *listen, Handler: service.Handler(), ReadHeaderTimeout: 5 * time.Second}
	serveErr := make(chan error, 1)
	go func() { serveErr <- server.ListenAndServe() }()
	ctx, stop := signal.NotifyContext(context.Background(), os.Interrupt, syscall.SIGTERM)
	defer stop()
	go api.RunProposalOutboxDispatcher(ctx, db, service)
	go api.RunResultOutboxDispatcher(ctx, db, service)
	go api.RunStateOutboxDispatcher(ctx, db, service)
	select {
	case err := <-serveErr:
		if err != nil && err != http.ErrServerClosed {
			fatalf("serve API: %v", err)
		}
	case <-ctx.Done():
		shutdownCtx, shutdownCancel := context.WithTimeout(context.Background(), 10*time.Second)
		defer shutdownCancel()
		if err := server.Shutdown(shutdownCtx); err != nil {
			fatalf("shutdown API: %v", err)
		}
	}
}

func newAPIHandler(db *sql.DB, workloadSecret string, indexes ...api.CandidateIndex) http.Handler {
	return newAPIService(db, workloadSecret, indexes...).Handler()
}

func newAPIService(db *sql.DB, workloadSecret string, indexes ...api.CandidateIndex) *api.Service {
	var candidateIndex api.CandidateIndex
	if len(indexes) > 0 {
		candidateIndex = indexes[0]
	}
	return &api.Service{
		SessionBackend:        store.PostgresSessions{DB: db},
		SessionIssuer:         store.PostgresSessions{DB: db},
		QueueBackend:          store.PostgresQueue{DB: db},
		ProposalBackend:       api.ProposalProviderFromStore(db),
		ProposalPromoter:      api.ProposalPromoterFromStore(db),
		ServerRegistrar:       api.ServerRegistrarFromStore(db),
		ServerShutdowner:      api.ServerShutdownerFromStore(db),
		ResultSubmitter:       store.PostgresResults{DB: db},
		RankedProfileProvider: store.PostgresRankedProfiles{DB: db},
		TierPolicy:            domain.DefaultTierPolicy(),
		Assignment:            api.AssignmentProviderFromStore(db),
		Roster: func(ctx context.Context, binding domain.WorkloadBinding, now time.Time) ([][]byte, error) {
			return store.GetAssignmentRoster(ctx, db, binding.MatchID, binding.ServerID, now)
		},
		CandidateIndex: candidateIndex,
		ProbeRecorder:  store.PostgresQueue{DB: db},
		WorkloadVerify: api.WorkloadVerifierFromSignedToken([]byte(workloadSecret), db),
		Now:            func() time.Time { return time.Now().UTC() },
		Log:            logEvent,
	}
}

// logEvent writes one credential-safe structured event per line to stderr.
// Best-effort: a logging failure must never fail or block the request it
// describes, so encode errors are swallowed rather than surfaced.
func logEvent(event observability.Event) {
	payload, err := observability.Encode(event)
	if err != nil {
		return
	}
	fmt.Fprintln(os.Stderr, string(payload))
}

func envOrDefault(name, fallback string) string {
	if value := os.Getenv(name); value != "" {
		return value
	}
	return fallback
}

func fatalf(format string, args ...any) {
	fmt.Fprintf(os.Stderr, "control-plane: "+format+"\n", args...)
	os.Exit(1)
}