Change nginx log directory owner to fix log rotation

[waj-kubiak]

Prior to this change, /var/log/nginx was owned by root. This resulted in a broken log rotation where nginx worker processes would continue to write to a rotated file access.log.1 with no further rotation until there's no disk space left. New log file access.log.0 would remain empty.

Changing the owner user to www-data so that it correlates with nginx and logrotate config files resolves the issue. Usually instances don't last long enough for this to be a problem, however it's been observed on few meter-points-service instances.

Tested with Vagrant, got no errors related to the change:
==> webhookserver: Notice: /Stage[main]/Octo_nginx/File[nginx log dir]/owner: owner changed 'root' to 'www-data'

Tested on an api-server instance on oegb-test after CIS hardening cron job kicks in:

[waj-kubiak]

Never mind this PR, this is likely the cause:

  # 5.4.4 Ensure default user umask is 027 or more restrictive
  file_line { "Ensure default user umask is 027 or more restrictive":
    path => "/etc/profile",
    line => "umask 027",
    match => "^umask "
  }