Restructure perfsonar-toolkit-security package

[arlake228]

It was decided we would like to further breakout the perfsonar-toolkit-security package instead more packages (maybe without toolkit in the name) that represent the individual functions so it is not all or nothing. This would give users more flexibility and make it easier for Ansible to manage firewalls. We probably ant packages for the following:

• open http(s)-related ports (i.e. esmond, pscheduler, maddash)
• open individual tool ports (OWAMP, iperf3, etc)
• Configure apache (and maybe an extra secure version for federal users)
• Configure memcached security
• Set up fail2ban

Will need to finalize how we organize these and may think of other things. We can still have a perfsonar-toolkit-security but it will depend on the sub-packages

[arlake228]

Look at adding a "-security" package to some of the individual packages that depend on these so people have an easy option to adding the security options they need.

[laeti-tia]

Once we have that, we should also then look if some of the sub-packages are worthy of use in the testpoint bundle (i.e. for securing httpd and memcached as these are used by the testpoint too).