Website Monitoring - Catching Loop Redirects

[Antonio Intagliata]

Hello, will Monit provide an option to catch loop redirects during a website page monitoring?

The configuration should have a maximum number of page redirects, after which the request is considered failed.

Thank you very much!

[Tildeslash]

Hello,

the http protocol test currently doesn't follow redirects, it could be interesting feature to add.

Best regards, The Monit team

[Henning Bopp]

I used the following "program" to monitor this for a single address:

curl --max-redirs 10 -Lvs -o/dev/null example.com 2>&1 | grep '^< Location:' | wc -l

Details:

• --max-redirs 10: Follow up to 10 redirects
• -L: Follow redirects
• -v: Verbose output (no output is generated otherwise)
• -s: silent: Do not print page bodies
• -o /dev/null: drop any output
• 2>&1: redirect stderr to stdout to be available with grep
• grep '^< Location:': Finds any line starting with < Location:
• wc -l: count lines

By using

check program RedirectsOnExampleCom with path "/usr/bin/bash -c 'exit $(curl --max-redirs 10 -Lvs -o/dev/null example.com 2>&1 | grep '^< Location:' | wc -l)'"

It will exit normally (exit code 0) if no redirect happened. It will walk up to 10 redirects, so one might to enhance like

check program RedirectsOnExampleCom with path "/usr/bin/bash -c 'exit $(curl --max-redirs 10 -Lvs -o/dev/null example.com 2>&1 | grep '^< Location:' | wc -l)'"
    if status > 2 then alert

So the alert is triggered if there are more than 2 redirects. That might make sence if you are redirected from http://example.com to http://www.example.com and finally to https://www.example.com.