向Grafana添加PostgreSQL数据源时插件健康检查失败

问题背景 #

使用Grafana Docker容器与PostgreSQL Docker容器，通过Python脚本调用Grafana HTTP API添加PostgreSQL数据源，已确认PostgreSQL用户为超级用户，但出现插件健康检查失败，执行查询时触发连接错误。

相关代码 #

datasource = {
            "name": "postgres",
            "type": "postgres",
            "host": f"http://{database.config.host}",
            "database": database.config.database,
            "user": database.config.user,
            "password": database.config.password,
            "access": "proxy",
            "port": database.config.port,
            "sslmode": "disable",
        }
headers = {
    "Content-Type": "application/json",
    "Authorization": f"Bearer {self.api_key}",
}
response = requests.post(
    self.grafana_url + self.API_DATASOURCES,
    json=datasource,
    headers=headers,
    timeout=2,
)

日志信息（中文翻译） #

logger=context t=2023-05-24T23:47:11.909546523Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:11.909856964Z level=warn msg=未授权 error="未找到用户令牌" remote_addr=192.168.16.1 traceID=
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:11.909908982Z level=info msg="请求完成" method=GET path=/api/live/ws status=401 remote_addr=192.168.16.1 time_ms=0 duration=790.303µs size=40 referer= handler=/api/live/ws
logger=context t=2023-05-24T23:47:24.906185083Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:24.906382401Z level=warn msg=未授权 error="未找到用户令牌" remote_addr=192.168.16.1 traceID=
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:24.906419291Z level=info msg="请求完成" method=GET path=/api/live/ws status=401 remote_addr=192.168.16.1 time_ms=0 duration=416.706µs size=40 referer= handler=/api/live/ws
logger=context t=2023-05-24T23:47:26.901251349Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:26.901763659Z level=warn msg=未授权 error="未找到用户令牌" remote_addr=192.168.16.1 traceID=
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:26.901862131Z level=info msg="请求完成" method=GET path=/api/live/ws status=401 remote_addr=192.168.16.1 time_ms=1 duration=1.014872ms size=40 referer= handler=/api/live/ws
logger=context t=2023-05-24T23:47:33.749570154Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:33.749850255Z level=info msg="请求完成" method=GET path=/connections/your-connections/datasources/edit/edec663f-a7cc-4316-9aa1-d672dee7d717 status=302 remote_addr=192.168.16.1 time_ms=0 duration=523.514µs size=29 referer=http://localhost:3000/login handler=/connections/your-connections/datasources/edit/*
logger=context t=2023-05-24T23:47:33.751853587Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=context t=2023-05-24T23:47:36.899061961Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:36.899246767Z level=warn msg=未授权 error="未找到用户令牌" remote_addr=192.168.16.1 traceID=
logger=context userId=0 orgId=0 uname= t=2023-05-24T23:47:36.899285702Z level=info msg="请求完成" method=GET path=/api/live/ws status=401 remote_addr=192.168.16.1 time_ms=0 duration=439.616µs size=40 referer= handler=/api/live/ws
logger=context t=2023-05-24T23:47:42.741638708Z level=warn msg="无法从Cookie中查找会话" error="未找到用户令牌"
logger=http.server t=2023-05-24T23:47:42.75135848Z level=info msg="登录成功" User=admin@localhost
logger=context userId=1 orgId=1 uname=newuser t=2023-05-24T23:47:42.894368364Z level=info msg="请求完成" method=GET path=/api/live/ws status=-1 remote_addr=192.168.16.1 time_ms=0 duration=841.72µs size=0 referer= handler=/api/live/ws
logger=context userId=1 orgId=1 uname=newuser t=2023-05-24T23:47:42.912548509Z level=info msg="请求完成" method=GET path=/api/datasources/uid/edec663f-a7cc-4316-9aa1-d672dee7d717 status=404 remote_addr=192.168.16.1 time_ms=0 duration=461.639µs size=35 referer=http://localhost:3000/connections/your-connections/datasources/edit/edec663f-a7cc-4316-9aa1-d672dee7d717 handler=/api/datasources/uid/:uid
logger=context userId=1 orgId=1 uname=newuser t=2023-05-24T23:47:42.946075661Z level=info msg="请求完成" method=GET path=/api/datasources/edec663f-a7cc-4316-9aa1-d672dee7d717 status=400 remote_addr=192.168.16.1 time_ms=0 duration=341.626µs size=27 referer=http://localhost:3000/connections/your-connections/datasources/edit/edec663f-a7cc-4316-9aa1-d672dee7d717 handler=/api/datasources/:id
logger=context userId=1 orgId=1 uname=newuser t=2023-05-24T23:47:56.90833634Z level=info msg="请求完成" method=GET path=/api/live/ws status=-1 remote_addr=192.168.16.1 time_ms=1 duration=1.658ms size=0 referer= handler=/api/live/ws
logger=context userId=1 orgId=1 uname=newuser t=2023-05-24T23:48:00.997970403Z level=error msg="插件健康检查失败" error="无法检查插件健康：健康检查失败" remote_addr=192.168.16.1 traceID=
logger=context userId=1 orgId=1 uname=newuser t=2023-05-24T23:48:00.998018152Z level=error msg="请求完成" method=GET path=/api/datasources/uid/fb4800be-7c42-40fb-be33-f3e0d87ebaab/health status=500 remote_addr=192.168.16.1 time_ms=0 duration=623.119µs size=53 referer=http://localhost:3000/connections/your-connections/datasources/edit/fb4800be-7c42-40fb-be33-f3e0d87ebaab handler=/api/datasources/uid/:uid/health

排查与解决方案 #

1. 修正数据源Host配置 #

PostgreSQL数据源的host字段不需要http://前缀，只需要填写容器名称或IP地址。修改代码：

"host": f"{database.config.host}",  # 移除http://前缀

2. 检查Docker容器网络连通性 #

• 确认Grafana与PostgreSQL容器在同一Docker网络中，若不在需添加到同一网络。
• 进入Grafana容器，执行ping <postgres容器名称/IP>测试网络可达性。

3. 验证PostgreSQL连接权限 #

• 修改PostgreSQL的pg_hba.conf文件，添加允许Grafana容器IP段连接的规则：

  host    目标数据库名    目标用户名    grafana容器IP/24    scram-sha-256
  

• 重启PostgreSQL容器使配置生效。

4. 确认端口参数类型 #

检查port字段是否为整数，若配置中读取的是字符串需转换：

"port": int(database.config.port),

5. 启用调试日志定位细节 #

设置Grafana容器环境变量GF_LOG_LEVEL=debug，重启后查看日志，获取更具体的连接失败原因（如密码错误、数据库不存在等）。

内容的提问来源于stack exchange，提问作者An old man in the sea.