alertmanager-webhookflask_Python_Dockerfile_下载.zip

## Alertmanager Webhook To store `resolved` status alerts in linux sqlite3 database by python flask framework. ### API #### alerts ['POST'] The interface used by Alertmanager webhook. ##### Configuration ```bash route: routes: - receiver: 'webhook' group_wait: 30s group_interval: 1m repeat_interval: 1h match_re: key: ${value} continue: true ... receivers - name: 'webhook' webhook_configs: - url: 'http://alertmanager-webhook:5000/alerts' send_resolved: true ``` #### alerts_history ['GET'] ##### Parameters | Name | type | default | | ------------ | -------- | ------------| | page | int | 1 | | per_page | int | 20 | | start | str | \ | | end | str | \ | | severity | str | \ | The value be returned order by start time. * Demo Request ```bash curl -g -i -X GET -H "Accept: application/json" "http://alertmanager-webhook:5000/alerts_history?page=1&per_page=20&start=1542963638&end=1542965018" ``` Response ```json { "alerts": [ { "alertname": "NodeCPUUsageOvercommit", "end": "Fri, 23 Nov 2018 10:24:42 GMT", "hash_id": "-2466042037021245317", "id": 1, "message": "High CPU Usage in Node", "resource": "192.168.0.2", "resource_type": "node", "severity": "critical", "start": "Fri, 23 Nov 2018 09:20:12 GMT" }, { ... }], "pagination": { "page": 1, "pages": 1, "per_page": 20, "total": 4 } } ``` ### ENV | Name | type | default | | ------------ | -------- | ------------| | PORT | int | 5000 | | DB_PATH | str | '/var/lib/alerts/alert.db' | | HOST | str | '0.0.0.0' | ### table `alerts` | Name | type | default | | ------------ | -------- | ------------| | id | int | autoincrement | | alertname | str | \ | | resource | str | \ | | message | text | \ | | hash_id | str | \ | | severity | str | \ | | start | datetime | datetime.datetime| | end | datetime | datetime.datetime | ### Troubleshooting #### alert uniqueness I found that `resolved` alerts will be sent repeatedly, so I used some return `values` to do hash. Like some alerts' info about `Node`, I used `alertname`， `instance`, `startsAt` this three value to do hash and achieve alert uniqueness. ```python def hash_value(labels, starts): hash_str = labels.get("alertname", '') + labels.get("instance", "") + starts return hash(hash_str) ``` ### Version `v1.0`