Files
pangolin/server/cmd/agent/main.go
T
wangjia f03d2dc8a6 feat(agent): node agent — enroll/mTLS, heartbeat, command stream, sing-box 用户表 (tsk__R8M4jEw43JR)
与控制面同仓同 go.mod,新增节点 agent 实现:

- proto/agent/v1/agent.proto + internal/pb/agentv1:冻结的控制面↔agent gRPC 契约
  (Enroll/Register/Heartbeat/Subscribe/Ack/ReportUsage)。仓库尚无 protoc 流水线,
  暂以手写 Go 类型 + JSON gRPC codec 实现,与 proto 1:1 对应,待 protoc 接入即可替换。
- internal/agentd:
  - enroll.go:首启生成 EC 密钥+CSR,持 bootstrap token 调 Enroll 换 90d 节点证书
    (CN=node_uuid),落 /etc/pangolin-agent/,此后 mTLS。
  - conn.go(agent.go)+creds.go:mTLS 主动拨号 + 指数退避重连;重连携带 last_command_id;
    Register 取 ConfigSnapshot 全量配置覆盖本地。
  - heartbeat.go:30s 上报 peer/带宽/CPU + config_version;need_full_resync→全量同步。
  - command.go:消费 Subscribe,Upsert/Revoke/Rotate/ApplyConfig/Lifecycle 幂等处理后
    Ack(at-least-once,按 command_id 去重)。
  - singbox.go+render.go:内存用户表 + 落盘 state.json(仅 dp_uuid+expires_at);任何变更
    渲染完整 sing-box 配置(REALITY users[uuid,flow] + Hy2 users[派生口令])→ 500ms 去抖
    合并 → systemd 重启。
  - ttl.go:凭证 TTL 定时移除并上报。
  - usage.go:按 dp_uuid 聚合上报,绝无 user_id/email/目的地址。
  - derive.go:Hy2 口令 = HMAC-SHA256(key, dp_uuid),与控制面同源派生。
- cmd/agent:入口(flag/env 配置)。
- infra/cloud-init/{node.yaml.tmpl,install-node.sh,README.md}:一段式安装,下载锁定版本
  二进制并校验 SHA-256,systemd 拉管,首启即 Enroll/Register。shellcheck -S warning 通过。

测试(bufconn mock 控制面,无需 docker):Enroll→Register→Heartbeat 全流转;Upsert/Revoke
渲染正确;Rotate 宽限期新旧并存到点移除;TTL 自动移除并上报;断流重连 last_command_id
续发不丢不重;need_full_resync 触发重注册;state.json 恢复;去抖合并;扫描确认无身份字段。
go test -race ./internal/agentd/... ./internal/pb/... 通过;go vet ./... 通过。

落实 doc/04 §2 节点无状态化与 doc/06 §3 数据面红线(节点仅见 dp_uuid)。

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-13 12:26:58 +08:00

59 lines
2.5 KiB
Go

// Command agent is the Pangolin node agent. It runs on every acceleration node,
// self-enrolls over mTLS, dials the control plane, applies the streamed command
// feed by managing the local sing-box user table, and reports heartbeat/usage.
//
// Configuration comes from flags or PANGOLIN_AGENT_* environment variables
// (cloud-init injects the latter). The node holds zero user identities on disk.
package main
import (
"context"
"flag"
"log"
"os"
"os/signal"
"syscall"
"time"
"github.com/wangjia/pangolin/server/internal/agentd"
)
func main() {
cfg := agentd.Config{}
var singboxUnit string
flag.StringVar(&cfg.ControlPlaneAddr, "control-plane", env("PANGOLIN_AGENT_CONTROL_PLANE", ""), "control plane gRPC address host:port")
flag.StringVar(&cfg.ServerName, "server-name", env("PANGOLIN_AGENT_SERVER_NAME", ""), "TLS SNI to validate (defaults to control-plane host)")
flag.StringVar(&cfg.BootstrapToken, "bootstrap-token", env("PANGOLIN_AGENT_BOOTSTRAP_TOKEN", ""), "one-time enrollment token")
flag.StringVar(&cfg.StateDir, "state-dir", env("PANGOLIN_AGENT_STATE_DIR", agentd.DefaultStateDir), "directory for node key/cert/ca/state")
flag.StringVar(&cfg.SingboxConfigPath, "singbox-config", env("PANGOLIN_AGENT_SINGBOX_CONFIG", agentd.DefaultSingboxCfg), "path to rendered sing-box config")
flag.StringVar(&cfg.DeriveKey, "derive-key", env("PANGOLIN_AGENT_DERIVE_KEY", ""), "Hy2 password derivation key")
flag.StringVar(&cfg.AgentVersion, "agent-version", env("PANGOLIN_AGENT_VERSION", "dev"), "reported agent version")
flag.StringVar(&singboxUnit, "singbox-unit", env("PANGOLIN_AGENT_SINGBOX_UNIT", "sing-box"), "systemd unit to restart on config change")
flag.BoolVar(&cfg.Insecure, "insecure", env("PANGOLIN_AGENT_INSECURE", "") == "1", "dial without mTLS (dev only)")
flag.Parse()
if cfg.ControlPlaneAddr == "" {
log.Fatal("agent: --control-plane (or PANGOLIN_AGENT_CONTROL_PLANE) is required")
}
agent := agentd.New(cfg, agentd.WithRestarter(agentd.SystemdRestarter{Unit: singboxUnit}))
ctx, stop := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM)
defer stop()
log.Printf("[pangolin-agent] starting (control-plane=%s, version=%s)", cfg.ControlPlaneAddr, cfg.AgentVersion)
start := time.Now()
if err := agent.Run(ctx); err != nil {
log.Fatalf("[pangolin-agent] exited after %s: %v", time.Since(start), err)
}
log.Printf("[pangolin-agent] shut down cleanly after %s", time.Since(start))
}
func env(key, def string) string {
if v, ok := os.LookupEnv(key); ok {
return v
}
return def
}