TiDB Ansible deploy error

[wolaiye1010]

到 启动这一步到时候报错 ansible-playbook start.yml，
错误内容如下：
TASK [wait up] **********************************************************************************************************************************************************************
fatal: [10.9.193.135]: FAILED! => {"changed": false, "elapsed": 300, "failed": true, "msg": "Timeout when waiting for PID:/home/tidb/deploy/status/tikv.pid to stop."}
fatal: [10.9.193.136]: FAILED! => {"changed": false, "elapsed": 300, "failed": true, "msg": "Timeout when waiting for PID:/home/tidb/deploy/status/tikv.pid to stop."}
fatal: [10.9.193.121]: FAILED! => {"changed": false, "elapsed": 300, "failed": true, "msg": "Timeout when waiting for PID:/home/tidb/deploy/status/tikv.pid to stop."}
to retry, use: --limit @retry_files/start.retry

PLAY RECAP **************************************************************************************************************************************************************************
10.9.193.119 : ok=23 changed=3 unreachable=0 failed=0
10.9.193.121 : ok=21 changed=3 unreachable=0 failed=1
10.9.193.135 : ok=19 changed=2 unreachable=0 failed=1
10.9.193.136 : ok=19 changed=2 unreachable=0 failed=1
localhost : ok=1 changed=0 unreachable=0 failed=0

我的配置文件：

TiDB Cluster Part

[tidb_servers]
10.9.193.119
10.9.193.121

[pd_servers]
10.9.193.119
10.9.193.121

[tikv_servers]
10.9.193.135
#10.9.193.137
10.9.193.121
10.9.193.136

Monitoring Part

[monitoring_servers]
10.9.193.119

[grafana_servers]
10.9.193.119

[monitored_servers:children]
tidb_servers
tikv_servers
pd_servers

Binlog Part

[pump_servers:children]
tidb_servers

[cistern_servers]

[drainer_servers]

[pd_servers:vars]

location_labels = ["zone","rack","host"]

Global variables

[all:vars]
deploy_dir = /home/tidb/deploy

Connection

ssh via root:

ansible_user = root
ansible_become = true
ansible_become_user = work

ssh via normal user

#ansible_user = tidb

cluster_name = test-cluster

misc

enable_elk = False
enable_firewalld = False
enable_ntpd = False
machine_benchmark = True
set_hostname = False

binlog trigger

enable_binlog = False

[buggithubs]

@wolaiye1010

0. 主机系统是否为ubuntu?
   1. ubunut 注释become与become_user；修改ansible_user = work
1. 检查文件权限/home/tidb/deploy/status/tikv.pid 是否为work
2. 可以尝试登录节点手动执行，然后查看是否有错误信息。
3. 以上均无误，可以尝试多次执行stop.yml & start.yml。

[wolaiye1010]

@buggithubs ,
主机系统为 CentOS 6.6
/home/tidb/deploy/status/tikv.pid 是work 权限
多次 stop.yml & start.yml 都不行，报同一个错误
测试主机不能链接外网，但我的localhost可以，我是通过 localhost 进行的ansible-playbook start.yml
不知道跟这个有没有关系，

“可以尝试登录节点手动执行，然后查看是否有错误信息。”
大神，请教下，怎么手动执行呢

[zimulala]

@LinuxGit

[LinuxGit]

@wolaiye1010 refer to https://github.com/pingcap/docs-cn/blob/master/op-guide/ansible-deployment.md. Please use CentOS 7 or later, default TiDB binaries is for CentOS7. And pd instances should be one or three.You can new issues in https://github.com/pingcap/tidb-ansible if you encounter any other deployment problems.

[wolaiye1010]

@LinuxGit ,
Think you, CentOS 7 is ok!