// Command dispatcher 启动 sundynix-dispatcher —— 第 4 层 AI Agent 调度集群。 package main import ( "context" "encoding/json" "log" "os" "os/signal" "syscall" "time" "github.com/sundynix/sundynix-dispatcher/internal/eino" "github.com/sundynix/sundynix-dispatcher/internal/harness" "github.com/sundynix/sundynix-dispatcher/internal/llm" dnats "github.com/sundynix/sundynix-dispatcher/internal/nats" "github.com/sundynix/sundynix-shared/otelx" "github.com/sundynix/sundynix-shared/secrets" ) func main() { secrets.MustHaveKeyInProd() // 生产须配 SUNDYNIX_SECRET_KEY 以解密下发的 api_key 密文 otelx.SetupSlog("sundynix-dispatcher") // 结构化日志 + 链路感知(任务日志带 trace_id) // 链路追踪:消费 span 续上 gateway 的 trace,再向下展开节点/工具/LLM span。 shutdownTrace, _ := otelx.Init(context.Background(), "sundynix-dispatcher") defer func() { _ = shutdownTrace(context.Background()) }() natsURL := envOr("NATS_URL", "nats://localhost:4222") pool := llm.NewPool() // LLM Pool: vLLM / Ollama 集群 breaker := harness.NewCircuitBreaker() // Harness: 熔断降级中心 // Harness: LLM 自动化评测(规则 + LLM-as-judge,模型就绪时启用)。 eval := harness.NewEvaluator(pool.Ready, func(ctx context.Context, sys, user string) (string, error) { return pool.Chat(ctx, []llm.ChatMessage{{Role: "system", Content: sys}, {Role: "user", Content: user}}) }) sub := dnats.MustConnect(natsURL) defer sub.Close() // 配置控制面:先订阅热更新,再后台重试拉初始配置(容忍 gateway 晚于本服务启动, // 避免一次性请求扑空后只能干等热更新 → 降级桩跑全程)。 if _, err := sub.SubscribeModelConfigUpdated(pool.SetConfig); err != nil { log.Printf("[dispatcher] subscribe model config: %v", err) } go sub.FetchModelConfigWithRetry(context.Background(), pool.SetConfig) // sub 同时作为 Token 回流(TokenSink)、MCP 工具调用(ToolCaller)、执行事件(ExecSink)与任务状态回写(StatusSink)出口。 orch, err := eino.NewOrchestrator(pool, breaker, eval, sub, sub, sub, sub) if err != nil { log.Fatalf("[dispatcher] build eino graph: %v", err) } // 健康心跳:dispatcher 无 HTTP/工具端点,挂一个 NATS 应答让管理端「服务状态」探到它在线。 startedAt := time.Now() if unsub, herr := sub.ServeHealth(func() []byte { data, _ := json.Marshal(map[string]any{ "ok": true, "service": "dispatcher", "model": pool.ModelName(), "ready": pool.Ready(), "uptime_s": int(time.Since(startedAt).Seconds()), }) return data }); herr != nil { log.Printf("[dispatcher] serve health: %v", herr) } else { defer func() { _ = unsub() }() } // 监听退出信号,优雅停止消费。 ctx, stop := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM) defer stop() log.Println("[dispatcher] consuming sundynix.tasks.* (Ctrl-C to quit)") if err := sub.ConsumeTasks(ctx, orch.Handle); err != nil && err != context.Canceled { log.Fatalf("[dispatcher] exit: %v", err) } } func envOr(key, def string) string { if v := os.Getenv(key); v != "" { return v } return def }