A tool to scrape multiple devices and push to prometheus gateway.
This application requires python installed.
Execute python3 -m pip install -r requirements.txt to install necessary requirements.
All configurations are written in json.
There will be a config_xxxx.json file for each type of device.
All targets have following template:
{
"conf": {
"prom_id": "ExporterINFO_PROJECT",
"interval": 55,
"timeout": 3,
"value_dev_down": 3000
},
"jobs": [
{
"job": "MyDevice01",
"ip": "192.168.1.57",
"metrics": [
"ping"
]
},
{
"job": "MyDevice01",
"ip": "192.168.1.57",
"metrics": [
"ping"
]
}
]
}Since we don't always have a connection between Prometheus server and targets, we would like to automatically retrieve some metrics and send them to PushGateway.
How PingExporter works?
- We define various configs project1_config.json files with above structure
- An ExporterEngine is given to this config and it will loop every jobs.
- A job is a targeted device with multiple metrics.
- An AbstractScraper is created for each metric (e.g. for "ping" is "PingScraper")
- Each scraper defines his own metrics (e.g. Pingscraper defines 'dev_ping')
Why do we need to specify multiple configs (e.g. project1_consig.json, project2_config.json)? Because we may want to monitor metrics every 30 seconds but i have 50 devices, so although ping is fast, we may break the 40 seconds barrier (we don't keep up with the export loop).
But if we split those devices in multiple configs (remember, each config has its own cycle in its own thread), we are ok. Besides that, we can define the scrape interval for each config!
Monitoring PingExporter performances is easy, since it exports itself a exporter_duration and exporter_job for each config.