Promethist Watchdog
Watching a server by its web address
Watchdog can watch a server it does not run, such as a signalling server on a cloud
machine. The server appears as a row in the table: green up while its address
answers, red down when it stops. You get an email when it goes down and another
when it comes back. Watchdog only looks — it never restarts anything on that server.
This is useful when the server cannot send email itself; Azure, for example, blocks outgoing mail from its virtual machines.
Setting it up
Click Add, give it a name, tick Only check a web address, and enter the address. The program settings do not apply, so they are hidden.
Or from the command line:
wdog app add signalling --url https://signalling.example.com/
When it counts as down
Watchdog opens the address every 30 seconds. Any answer below HTTP 400 counts as
working. An error such as 502 Bad Gateway, or no answer within 10 seconds, is a
failed check. After two failed checks in a row, the row turns red and the
url_down email goes out. When the address answers again, the url_restored email
says how long it was down.
All three numbers are in the app’s settings. Hover over the status to see the last answer, or why there was none.
url_down and url_restored are on by default. See
Email alerts.