ansible sudo_user挂起几分钟然后失败(在centos6.5.1 vagrant vm中)

我有这两个简单的任务:

- name: I am 
  shell: "echo `id`"

- name: say hello
  shell: echo "postgres saying hello"
  sudo_user: postgres

第二个任务在经过一段长时间的停顿后失败,输出低于(它正在使用详细级别的vvv运行)(是的,我已经验证用户postgres存在,我可以从VM内部执行sudo su postgres)

TASK: [postgresql | I am] ***************************************************** 
changed: [192.168.78.6] => {"changed": true, "cmd": "echo `id` ", "delta": "0:00:00.002511", "end": "2014-01-23 22:49:14.161249", "item": "", "rc": 0, "start": "2014-01-23 22:49:14.158738", "stderr": "", "stdout": "uid=0(root) gid=0(root) groups=0(root)"}

TASK: [postgresql | say hello] ************************************************ 
fatal: [192.168.78.6] => failed to parse: [sudo via ansible, key=fnfgfnxabemrzbfixwgoksvgjrfzplxf] password: 


FATAL: all hosts have already failed -- aborting

东西在centos6.5.1 vagrant vm中运行