prometheus Alertmanager报错

[sammasamadhi]

使用kubeasz-0.5.1部署成功后,在部署prometheus的时候 alertmanager组件报错如下:

 level=info ts=2019-01-24T09:20:01.118150229Z caller=main.go:174 msg="Starting Alertmanager" version="(version=0.15.2, branch=HEAD, revision=d19fae3bae451940b8470abb680cfdd59bfa7cfa)"
level=info ts=2019-01-24T09:20:01.118241804Z caller=main.go:175 build_context="(go=go1.10.3, user=root@3101e5b68a55, date=20180814-10:53:39)"
level=warn ts=2019-01-24T09:20:01.122920737Z caller=cluster.go:148 component=cluster err="couldn't deduce an advertise address: no private IP found, explicit advertise addr not provided"
level=error ts=2019-01-24T09:20:01.129140557Z caller=main.go:199 msg="unable to initialize gossip mesh" err="create memberlist: Failed to get final advertise address: No private IP address found, and explicit IP not provided" 

[sammasamadhi]

把alertmanager版本降低到prom/alertmanager:v0.14.0 就没有问题

[gjmzj]

我本地有好几个实例配置的就是 v0.15.2，都没碰到这个问题，尝试去 prometheus 项目那边反馈看看

[weilinqwe]

我也碰到了
`

[root@cloudhost1 prometheus]# kubectl get pod   -n monitoring                                                  
NAME                                                     READY     STATUS             RESTARTS   AGE
grafana-796ffb9558-4d4fw                                 1/1       Running            0          17m
monitor-prometheus-alertmanager-67dd877d66-xxdlj         1/2       CrashLoopBackOff   1          18s
monitor-prometheus-kube-state-metrics-78b6c5f4d9-fk2k7   1/1       Running            0          18m
monitor-prometheus-node-exporter-brk52                   1/1       Running            0          18m
monitor-prometheus-node-exporter-h8dgb                   1/1       Running            0          18m
monitor-prometheus-node-exporter-mlx88                   1/1       Running            0          18m
monitor-prometheus-node-exporter-xgznb                   1/1       Running            0          18m
monitor-prometheus-node-exporter-zc2hq                   1/1       Running            0          18m
monitor-prometheus-server-8f44b6c7d-hdtl4                2/2       Running            0          18m
[root@cloudhost1 prometheus]# kubectl logs  monitor-prometheus-alertmanager-67dd877d66-xxdlj -c prometheus-alertmanager-configmap-reload -n monitoring
2019/03/14 12:18:18 Watching directory: "/etc/config"
[root@cloudhost1 prometheus]# kubectl logs  monitor-prometheus-alertmanager-67dd877d66-xxdlj -c prometheus-alertmanager -n monitoring
level=info ts=2019-03-14T12:19:58.378674034Z caller=main.go:174 msg="Starting Alertmanager" version="(version=0.15.2, branch=HEAD, revision=d19fae3bae451940b8470abb680cfdd59bfa7cfa)"
level=info ts=2019-03-14T12:19:58.378827621Z caller=main.go:175 build_context="(go=go1.10.3, user=root@3101e5b68a55, date=20180814-10:53:39)"
level=warn ts=2019-03-14T12:19:58.394107651Z caller=cluster.go:148 component=cluster err="couldn't deduce an advertise address: no private IP found, explicit advertise addr not provided"
level=error ts=2019-03-14T12:19:58.421706704Z caller=main.go:199 msg="unable to initialize gossip mesh" err="create memberlist: Failed to get final advertise address: No private IP address found, and explicit IP not provided"

`
一开始我怀疑双网卡，default gw在外网卡的缘故，改了之后没用，再研究一下

[weilinqwe]

@gjmzj

我搜了一下，是prometheus的charts bug好像。
helm/charts#6507
有可能新版本已经修复了。

我按照上面的解决方案，修改了charts中的alarmmanager-deployment.yaml文件， 重装了prometheus charts就可以了。

估计群主prometheus charts版本太低（7.1.4）了，或者从helm仓库取最新的装，或者按照上述修改一下
这个大概需要在文档中标注一下。

[3CB3EC]

启动的时候增加参数--cluster.advertise-address=0.0.0.0:9093（或者改成你想要的端口）
解决。。

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.