I would like to also monitor non-crucial services with nagios like for example our GitLab-server or phpMyAdmin instance. Is there any way to just create warnings instead of circuital errors for some services?

At the moment I'm using the following:

define service {
    host_name localhost
    use generic-service
    service_description HTTP GitLab
    check_command check_www!git.example.com!'/users/sign_in'
}

define command {
    command_name check_www
    command_line /usr/lib/nagios/plugins/check_http -H '$ARG1$' -I '$HOSTADDRESS$' -e 'HTTP/1.1 200 OK' -u '$ARG2$'
}