Fix (target start): curveadm target start timeout

Signed-off-by: zztaki <[email protected]>

Author: zztaki

internal/task/step/container.go

@@ -119,6 +119,7 @@ type (
 	ContainerExec struct {
 		ContainerId *string
 		Command     string
+		Detached    bool
 		Success     *bool
 		Out         *string
 		module.ExecOptions
@@ -289,6 +290,9 @@ func (s *ListContainers) Execute(ctx *context.Context) error {
 
 func (s *ContainerExec) Execute(ctx *context.Context) error {
 	cli := ctx.Module().DockerCli().ContainerExec(*s.ContainerId, s.Command)
+	if s.Detached {
+		cli.AddOption("--detach")
+	}
 	out, err := cli.Execute(s.ExecOptions)
 	return PostHandle(s.Success, s.Out, out, err, errno.ERR_RUN_COMMAND_IN_CONTAINER_FAILED.FD("(%s exec CONTAINER COMMAND)", s.ExecWithEngine))
 }

internal/task/task/bs/start_tgtd.go

@@ -127,7 +127,8 @@ func NewStartTargetDaemonTask(curveadm *cli.CurveAdm, cc *configure.ClientConfig
 		ExecOptions: curveadm.ExecOptions(),
 	})
 	t.AddStep(&step.ContainerExec{
-		Command:     "tgtd -f &",
+		Command:     "tgtd -f",
+		Detached:    true,
 		ContainerId: &containerId,
 		ExecOptions: curveadm.ExecOptions(),
 	})