OpsGenie `api_host` incorrectly documented

[azelezni]

What did you do?

1. Configured Alertmanager with OpsGenie receiver
2. Created an alert in Grafana UI that will always fire

What did you expect to see? Expected Alertmanager to create a new alert in OpsGenie

What did you see instead? Under which circumstances? No alert created in OpsGenie, looking at the logs I get this error:

Notify for alerts failed" num_alerts=1 err="devops-opsgenie/opsgenie[0]: notify retry canceled due to unrecoverable error after 1 attempts: unexpected status code 405: {\"message\":\"Request method 'POST' not supported\",\"took\":0.001,\"requestId\":\"d50741e2-e518-4d31-b562-7b1d0ac21f4d\"}

Environment

• System information:

  Linux 5.4.219-126.411.amzn2.x86_64 x86_64

• Alertmanager version:

  alertmanager, version 0.24.0 (branch: HEAD, revision: f484b17fa3c583ed1b2c8bbcec20ba1db2aa5f11)
  build user:       root@265f14f5c6fc                                                          
  build date:       20220325-09:31:33                                                          
  go version:       go1.17.8                                                                   
  platform:         linux/amd64  

• Prometheus version:

  prometheus, version 2.40.5 (branch: HEAD, revision: 44af4716c86138869aa621737139e6dacf0e2550)
  build user:       root@70f803b28803                                                        
  build date:       20221201-12:50:06                                                        
  go version:       go1.19.3                                                                 
  platform:         linux/amd64  

• Alertmanager configuration file:

  global:
  resolve_timeout: 5m
  opsgenie_api_url: https://api.opsgenie.com/v2/alerts
  receivers:
  - name: "null"
  - name: devops-opsgenie
  opsgenie_configs:
    - api_key_file: /opt/opsgenie/opsgenie_api_key_devops
      description: test
      message: test
      priority: '{{ range .Alerts }}{{ if eq .Labels.severity "critical"}}P1{{else}}P3{{end}}{{end}}'
      responders:
        - name: DevOps
          type: team
      send_resolved: true
      tags: '{{ range .Alerts }}{{ .Labels.Values | join "," }}{{end}}'
  route:
  group_by:
  - grafana_folder
  - alertname
  group_interval: 5m
  group_wait: 30s
  receiver: "null"
  repeat_interval: 12h
  routes:
  - group_by:
      - '...'
    group_interval: 30s
    group_wait: 1s
    matchers:
      - team = "devops"
    receiver: devops-opsgenie
    repeat_interval: 30m
  templates:
  - /etc/alertmanager/config/*.tmpl

Now, I was able to solve this by simply removing the key global.opsgenie_api_url, and upon reading the docs again I noticed the strange wording used:

# The host to send OpsGenie API requests to.
[ api_url: [<string>](https://prometheus.io/docs/alerting/latest/configuration/#string) | default = global.opsgenie_api_url ]

So which is it? host or url? this is a big distinction and the above just makes it confusing. The default value configured at global.opsgenie_api_url is https://api.opsgenie.com/, this is a URL.

[simonpasquier]

:wave: why did you change the default opsgenie_api_url value? It should be the base URL of the OpsGenie API (e.g. https://api.opsgenie.com/ or https://api.eu.opsgenie.com/).

[azelezni]

Well that's why I'm here isn't it, the documentation isn't clear at all about the actual value. The doc says: The host to send OpsGenie API requests to The key is: api_url but the actual expected value is the root URL?

[simonpasquier]

but the actual expected value is the root URL?

yes, feel free to send a PR to improve the documentation :)