Files
backend/internal/grpc/runner/server.go
lan 8aa85ca4b2
All checks were successful
Build Backend / build (push) Successful in 2m22s
Build Backend / build-docker (push) Successful in 1m20s
feat(runner): implement cluster mode with Redis-based task dispatching and registry
Introduce a distributed task execution system for gRPC runners. This includes:
- A `TaskBus` for cluster-mode task dispatching via Redis Pub/Sub.
- A `RunnerRegistry` to track active runner instances in Redis.
- Support for both standalone and cluster modes via configuration.
- Refactored `RunnerHub` and `TaskManager` to use a `TaskDispatcher` interface, decoupling task submission from local execution.
- Added distributed locking to `HotRankWorker` to ensure only one instance performs periodic hot rank recalculations in a multi-instance deployment.
- Updated dependency injection (Wire) to support the new runner components and registry.
2026-05-07 01:08:39 +08:00

154 lines
3.4 KiB
Go
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
package runner
import (
"context"
"fmt"
"net"
"time"
"with_you/internal/config"
"with_you/proto/runner"
"go.uber.org/zap"
"google.golang.org/grpc"
"google.golang.org/grpc/credentials"
"google.golang.org/grpc/credentials/insecure"
)
// Server gRPC 服务器
type Server struct {
cfg *config.GRPCConfig
logger *zap.Logger
hub *RunnerHub
grpcSrv *grpc.Server
}
// NewServer 创建新的 gRPC 服务器(内部构造 Hub + TaskManager
func NewServer(cfg *config.GRPCConfig, logger *zap.Logger) *Server {
taskManager := NewTaskManager(nil, 60*time.Second, logger)
hub := NewRunnerHub(taskManager, logger)
taskManager.SetDispatcher(hub)
return &Server{
cfg: cfg,
logger: logger,
hub: hub,
}
}
// NewServerWithDeps 使用外部依赖创建 gRPC 服务器Wire DI 使用)
func NewServerWithDeps(cfg *config.GRPCConfig, logger *zap.Logger, hub *RunnerHub) *Server {
return &Server{
cfg: cfg,
logger: logger,
hub: hub,
}
}
// Start 启动 gRPC 服务器
func (s *Server) Start() error {
if !s.cfg.Enabled {
s.logger.Info("gRPC server is disabled")
return nil
}
lis, err := net.Listen("tcp", s.cfg.Address())
if err != nil {
return fmt.Errorf("failed to listen on %s: %w", s.cfg.Address(), err)
}
// 配置 gRPC 服务器选项
var opts []grpc.ServerOption
// 配置 TLS
if s.cfg.TLSEnabled {
if s.cfg.TLSCertFile == "" || s.cfg.TLSKeyFile == "" {
return fmt.Errorf("TLS enabled but cert/key files not specified")
}
creds, err := credentials.NewServerTLSFromFile(s.cfg.TLSCertFile, s.cfg.TLSKeyFile)
if err != nil {
return fmt.Errorf("failed to load TLS credentials: %w", err)
}
opts = append(opts, grpc.Creds(creds))
s.logger.Info("gRPC server TLS enabled")
} else {
s.logger.Warn("gRPC server running without TLS")
}
s.grpcSrv = grpc.NewServer(opts...)
runner.RegisterRunnerHubServer(s.grpcSrv, s.hub)
s.logger.Info("gRPC server starting", zap.String("address", s.cfg.Address()))
go func() {
if err := s.grpcSrv.Serve(lis); err != nil {
s.logger.Error("gRPC server error", zap.Error(err))
}
}()
// 启动后台任务清理协程
go s.backgroundCleanup()
return nil
}
// backgroundCleanup 后台清理任务
func (s *Server) backgroundCleanup() {
ticker := time.NewTicker(30 * time.Second)
defer ticker.Stop()
for range ticker.C {
// 清理超时的 Runner 连接
s.hub.CheckStaleRunners(90 * time.Second)
}
}
// Stop 停止 gRPC 服务器
func (s *Server) Stop() {
if s.grpcSrv != nil {
s.logger.Info("Stopping gRPC server")
s.grpcSrv.GracefulStop()
}
}
// GetHub 获取 RunnerHub
func (s *Server) GetHub() *RunnerHub {
return s.hub
}
// GetTaskManager 获取 TaskManager
func (s *Server) GetTaskManager() *TaskManager {
return s.hub.taskManager
}
// Shutdown 优雅关闭
func (s *Server) Shutdown(ctx context.Context) error {
if s.grpcSrv == nil {
return nil
}
done := make(chan struct{})
go func() {
s.grpcSrv.GracefulStop()
close(done)
}()
select {
case <-done:
s.logger.Info("gRPC server stopped gracefully")
return nil
case <-ctx.Done():
s.logger.Warn("gRPC server shutdown timeout, forcing stop")
s.grpcSrv.Stop()
return ctx.Err()
}
}
// GetClientCredentials 获取客户端凭证配置
func GetClientCredentials(cfg *config.GRPCConfig) (credentials.TransportCredentials, error) {
if cfg.TLSEnabled {
return credentials.NewClientTLSFromFile(cfg.TLSCertFile, "")
}
return insecure.NewCredentials(), nil
}