Port Timeout times make no sense

Issue #115 resolved
Josh Goldsmith
created an issue

The port timeout numbers I'm seeing make no sense to me.

From the monit 5.10 web GUI:

Port If failed [10.1.2.104:3156/status.status56 [HTTP via TCP] with timeout 20000 seconds] for 5 cycles then restart Port If failed [10.1.2.104:3056 [DEFAULT via TCP] with timeout 5000 seconds] then restart

The process description:

check process php-fpm-php5.6 with pidfile /var/run/php-fpm-php5.6.pid stop program = "/sbin/service php-fpm-php5.6 stop" start program = "/sbin/service php-fpm-php5.6 start" if failed host 10.1.2.104 port 3056 type TCP then restart if failed host 10.1.2.104 port 3156 protocol http and request '/ping.status56' with timeout 20 seconds for 5 cycles then restart

Is there a divide by 1000 error in the Port Timeout logic?

Comments (5)

  1. Vlad Glagolev

    Hello,

    is it included with 5.11? because it still doesn't make any sense (though cli now reports correct values of seconds -- in comparison with 5.10).

    I'm trying to set up the following:

    check host a with address b.tld
        if failed ping count 5 with timeout 10 seconds
            then alert
    

    and it doesn't work, but once I remove extra attributes:

    check host a with address b.tld
        if failed ping
            then alert
    

    it starts to work, AND syslog shows the following insane values:

    Jan 05 22:55:38 [monit] Ping response for b.tld 1/3 timed out -- no response within 5000 seconds_
    Jan 05 22:55:48 [monit] Ping response for b.tld 2/3 timed out -- no response within 5000 seconds_
    Jan 05 22:55:58 [monit] Ping response for b.tld 3/3 timed out -- no response within 5000 seconds_
    Jan 05 22:55:58 [monit] 'a' ping test failed_
    

    but monit -t -c ... -v reports normal values of seconds for timeout in both cases:

    • Ping = if failed [count 3 with timeout 5 seconds] then alert -- default value
    • Ping = if failed [count 5 with timeout 10 seconds] then alert
  2. Log in to comment