Experimental YAML-based service monitoring thingy.
Via go:
go get github.com/sosedoff/howdy
Or download a binary release:
wget -O /usr/local/bin/howdy RELEASE_URL
chmod +x /usr/local/bin/howdy
All stuff goes into config:
name: myapp
enabled: true
notify:
slack:
webhook: "https://hooks.slack.com/services/..."
# channel is optional
channel: "#myapp"
checks:
# check live urls
web:
# simple url
- url: http://myapp.com/
# specify expected http status code
- url: http://myapp.com/login
status: 403
# specify code and format
- url: http://api.myapp.com/v1/foobar
status: 200 # default
format: json
# formats are: html, json, xml
# or it could be anything that's included into Content-Type response header
# check if host is reachable
ping:
- host: myapp.com
# check if domains are resolving correctly
dns:
- domain: myapp.com
ip: 1.2.3.4
# check if port is reachable. default timeout is 5 seconds
port:
- host: myapp.com
port: 22
- host: 127.0.0.1
port: 5000
net: udp # default is tcpSee all available options:
$ howdy --help
Usage of ./howdy:
-c string
Path to all configs
-d string
Path to database (default "howdy.sqlite3")
-l string
Path to log
-n Send notifications (default true)
-s Start server
-t Test mode
-v Show version
Run checks on all configs:
$ howdy -c /path/to/configs
Or run checks on a single config:
$ howdy /path/to/config.yml
- Add support for notifications via email
- Add ability to detect service recovery. That will require having a DB.
MIT