fix 1900 failures of the following case issue:
`name[casing]: All names should start with an uppercase letter.`
Signed-off-by: Ryan Lerch <rlerch@redhat.com>
For synchronizing from host A to host B we would have to have SSH
configured from A to B or from B to A (which we don't, and we don't even
want to). The slurp solution would work, but certbot is pretty picky
WRT the restored /etc/letsencrypt directory (structure, filenames,
symlinks..).
We need two-phase configuration for Let's Encrypt:
1. initialize; the letsencrypt automation, this is happening only
once per VM (when it is spawned)
2. periodic; using certbot-renew.timer.
Both those phases are now wrapped into copr/certbot role.
Phase 1. needs to be done before the web-server is started (so
include the role on appropriate place in the playbook), so we can
do 'certbot --standalone' (bounds to port 80).
Phase 2. is accomplished using the running web-server, using
'certbot renew --webroot'.
If (and only if) the certificate is renewed, web server needs to
be restarted (and lighttpd needs to have post-processed
certificate format). So we also need the deploy hook script in
hand (two actually, lighttpd/httpd) so 'certbot renew'
automatically does what we expect it to do.
