Ssl – nagios check_http does not produce right answer for some ssl sites

Are you sure you are calling the same plugin with the same arguments? It seems to me that you did not define the service correctly. Usually, the arguments are passed without options such as:

check_command   check_http!blablabla!follow

They are spearated by !. In the command definition, you can specify the correct options of the check_http plugin such as:

# 'check_http' command definition
define command{
        command_name    check_http
        command_line    /usr/lib/nagios/plugins/check_http -H '$HOSTADDRESS$' -s '$ARG1$' -f '$ARG2$'
        }

The arguments $ARG1$ and $ARG2$ will be replaced with the real values you specified in the service definition.

Host checks and service checks are almost entirely unrelated, except for an implicit dependency of the service on its associated host.

You have configured the host check for your custom time period, but the service check is using whichever check_period is defined in the template it's using. Add a check_period to your service definition to fix this.

Alternatively, you could use your custom time period as a notification_period for the service, if you just want to suppress notifications during the backup.

Also, not that exclusions might be broken in 3.3.x, as noted in this line from the Nagios version history under 3.2.0:

Known issue: Service checks that are defined with timeperiods that contain "exclude" directives are incorrectly re-scheduled. Don't use these for now - we'll get this fixed for 3.4

... so you might want to upgrade to 3.4.x or 3.5.0 (latest as of this writing).