Skip to content

Latest commit

 

History

History
103 lines (76 loc) · 7.17 KB

export_http.md

File metadata and controls

103 lines (76 loc) · 7.17 KB
Raw

Exporting data via HTTP transport

All methods allow data to be can be filtered by tags. So, if you have configuration as defined here https://github.com/vodolaz095/dashboard/blob/master/contrib/dashboard.yaml and dashboard running on http://localhost:3000, you can get all metrics via URL like this:

http://localhost:3000/

and you can filter readings only to mysql and postgres database related via URL like this:

http://localhost:3000/?dialect=sql&kind=database

Method 1. Via Browser

Just open http://localhost:3000 via browser to see something like this

dashboard_example.png elinks.png mobile.jpg

Data can be filtered by tags. If you want TLS encryption or even basic authorization, you can do it by reverse proxy with config like this one https://github.com/vodolaz095/dashboard/blob/master/contrib/nginx/dashboard.conf

It is worth notice data is updated in realtime using Server Sent Events technology

Filtering sensors by tags is supported, so this should work with default config http://localhost:3000/?kind=database&dialect=sql

Method 2. Via JSON endpoint

You can receive all dashboard data in JSON format by calling appropriate endpoint http://localhost:3000/json

Filtering sensors by tags is supported, so this should work with default config http://localhost:3000/json?kind=database&dialect=sql

Method 3: Via HTTP REST API

You can receive all dashboard sensors' reading in JSON format by calling appropriate endpoint http://localhost:3000/api/v1/sensor

Filtering sensors by tags is supported, so this should work with default config http://localhost:3000/api/v1/sensor?a=b

You can load particular sensors readings via this call using its name, for example, for load1 sensor readings can be retrieved via http://localhost:3000/api/v1/sensor/load1

Method 4. Via Prometheus v4 Metrics scrapper endpoint

Endpoint GET /metrics exposes sensor readings in Prometheous v4 format.

Filtering sensors by tags is supported, so this should work with default config http://localhost:3000/metrics?kind=database&dialect=sql

Method 4. Via plain text (console+curl friendly metrics)

Endpoint GET /text exposes sensor readings in plain text format. Filtering sensors by tags is supported, so this should work with default config http://localhost:3000/text?kind=database&dialect=sql

Method 5. Via CSV metrics

Endpoint /csv exposes sensor readings in plain text format with CSV encoding, something like this:

Name Minimum Value Maximum Error UpdatedAt Description
load1 0.0000 0.4700 0.0000 00:43:04 Get system load average during last minute https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
load5 0.0000 0.4700 0.0000 00:43:04 Get system load average during last 5 minutes https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
load15 0.0000 0.4700 0.0000 00:43:05 Get system load average during last 15 minutes https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
process 0.0000 1204.0000 0.0000 00:43:05 Number of currently running processes https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
free_ram 500.0000 3462.0469 8000.0000 00:43:05 Current free Random Access Memory volume in megabytes https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
free_home 0.0000 10620.6289 0.0000 00:43:06 Size of free space in megabytes in /home directory https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
used_home 0.0000 109758.9336 0.0000 00:43:06 Size of used disk space in megabytes in /home directory https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
ratio_home 0.0000 91.1774 0.0000 00:43:06 Free space ratio for /home directory in percents https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_linux_system.md
echo 0.0000 6.0000 60.0000 00:43:06 Get current second executing date command https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_shell.md
endpoint 0.0000 0.0000 0.0000 00:00:00 Update value by incoming POST request https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_endpoint.md
redis 0.0000 0.0000 0.0000 redis: nil 00:43:06 Get value of redis key a https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_redis.md
mysql 1.0000 87.4750 100.0000 00:43:06 Select random number from range https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_sql.md
postgres 1.0000 5.6523 100.0000 00:43:07 Select random number from range https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_sql.md
thermal0 1.0000 44.0000 100.0000 00:43:07 Get thermal sensor status from area 0 https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_file.md
redis subscriber 1.0000 0.0000 100.0000 00:00:00 Subscribe to redis channel and get values from it https://github.com/vodolaz095/dashboard/blob/master/docs/sensor_redis.md

Filtering sensors by tags is supported, so this should work with default config http://localhost:3000/csv?kind=database&dialect=sql

Extra methods

  • GET /healthcheck allows to check if all remote database connections are pingable and service is alive. In general, it is not save to expose this endpoint to all internet, so it can be concealed by nginx config like these dashboard.conf
  • GET /ping returns http status of 200 (OK) and Pong as body to show service is working. Endpoint is save to expose to unauthenticated users - it does nothing dangerous.