1 of my shared hosting servers running Plesk 12.5 was hacked from 1 vhost.
Scripts were uploaded into a compromised CMS, which created symlinks to config-files of all other vhosts on the server.
These symlinks were created using CGI/Perl. All files of other users were readable.
All this was made possible because of a simple .htaccess which enabled CGI and FollowSymLinks.
Right now I think I tackled the issue by configuring nginx to allow symlinks only to files of the same users.
But straight out of the box Plesk seems to be vulnerable to this.
Any thoughts on this?
Scripts were uploaded into a compromised CMS, which created symlinks to config-files of all other vhosts on the server.
These symlinks were created using CGI/Perl. All files of other users were readable.
All this was made possible because of a simple .htaccess which enabled CGI and FollowSymLinks.
Right now I think I tackled the issue by configuring nginx to allow symlinks only to files of the same users.
But straight out of the box Plesk seems to be vulnerable to this.
Any thoughts on this?